WO2015198767A1 - 異常検知装置及び異常検知方法 - Google Patents
異常検知装置及び異常検知方法 Download PDFInfo
- Publication number
- WO2015198767A1 WO2015198767A1 PCT/JP2015/064866 JP2015064866W WO2015198767A1 WO 2015198767 A1 WO2015198767 A1 WO 2015198767A1 JP 2015064866 W JP2015064866 W JP 2015064866W WO 2015198767 A1 WO2015198767 A1 WO 2015198767A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- state
- change
- person
- image
- persons
- Prior art date
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/20—Movements or behaviour, e.g. gesture recognition
- G06V40/23—Recognition of whole body movements, e.g. for sport training
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/60—Static or dynamic means for assisting the user to position a body part for biometric acquisition
- G06V40/63—Static or dynamic means for assisting the user to position a body part for biometric acquisition by static guides
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B13/00—Burglar, theft or intruder alarms
- G08B13/18—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength
- G08B13/189—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems
- G08B13/194—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems
- G08B13/196—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems using television cameras
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B13/00—Burglar, theft or intruder alarms
- G08B13/18—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength
- G08B13/189—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems
- G08B13/194—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems
- G08B13/196—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems using television cameras
- G08B13/19602—Image analysis to detect motion of the intruder, e.g. by frame subtraction
- G08B13/19613—Recognition of a predetermined image pattern or behaviour pattern indicating theft or intrusion
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B25/00—Alarm systems in which the location of the alarm condition is signalled to a central station, e.g. fire or police telegraphic systems
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
- H04N23/60—Control of cameras or camera modules
- H04N23/61—Control of cameras or camera modules based on recognised objects
- H04N23/611—Control of cameras or camera modules based on recognised objects where the recognised objects include parts of the human body
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
- H04N23/60—Control of cameras or camera modules
- H04N23/64—Computer-aided capture of images, e.g. transfer from script file into camera, check of taken image quality, advice or proposal for image composition or decision on when to take image
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
- H04N23/60—Control of cameras or camera modules
- H04N23/69—Control of means for changing angle of the field of view, e.g. optical zoom objectives or electronic zooming
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N7/00—Television systems
- H04N7/18—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast
Definitions
- the present invention relates to a technique for detecting an abnormality using an image.
- Patent Document 1 proposes a monitoring image display system that can confirm the presence or absence of an abandoned object or a lost object. Specifically, this system extracts, from an image obtained by capturing a predetermined region in time series, an image region that is different from the reference image of the predetermined region and whose position does not change with time. .
- Patent Document 2 proposes a video monitoring apparatus that automatically detects a suspicious person having a specific intention among persons appearing in an image.
- This apparatus extracts information representing a person's interest and psychological state from the person's gaze direction information, and uses the extracted information to search for a person in a specific psychological state. Specifically, this apparatus detects a person from the captured image, extracts the gaze direction information of the person, and gaze characteristic amount (specific direction total gaze time, number of gaze transitions, etc.) from the gaze direction information for each person ) And information on the behavior of the person imaged from the gaze feature amount is acquired.
- Patent Document 1 detects the presence of a left object that is not normally present in a photographed image as an abnormal state of the field.
- the apparatus of Patent Document 2 detects a person in a specific psychological state in a photographed image as a person abnormal state.
- various abnormal conditions can occur at the monitoring location.
- various abnormal states may occur in the park, such as the presence of abandoned items, filth, suspicious persons, breakage of playground equipment, occurrence of accidents, and the like.
- Each of the above-described proposed methods cannot comprehensively detect such multiple types of abnormal states.
- the present invention has been made in view of such circumstances, and it is an object of the present invention to provide a technique for comprehensively detecting various abnormal states.
- the abnormal state to be detected is not limited.
- the first aspect relates to an image processing apparatus.
- An image processing apparatus detects change in a state of a target in a person based on an input image, detects that a change in the state of the target has occurred in a plurality of people, and responds to the detection Determining means for determining an abnormal state.
- the second aspect relates to an abnormality detection method executed by at least one computer.
- the abnormality detection method according to the second aspect detects a change in the state of a target in a person based on an input image, detects that the change in the state of the target has occurred in a plurality of people, and determines an abnormal state according to the detection. Including judging.
- a program for causing at least one computer to execute the method of the second aspect or a computer-readable recording medium recording such a program. May be.
- This recording medium includes a non-transitory tangible medium.
- FIG. 4 is a flowchart illustrating an operation example of the image processing apparatus according to the first embodiment. It is a figure which shows notionally the process structural example of the image processing apparatus in 2nd embodiment. It is a flowchart which shows the operation example of the image processing apparatus in 2nd embodiment. It is a figure which shows notionally the process structural example of the image processing apparatus in 3rd embodiment. 10 is a flowchart illustrating an operation example of an image processing apparatus according to a third embodiment.
- FIG. 1 is a diagram conceptually illustrating a hardware configuration example of an image processing apparatus 1 (hereinafter also simply referred to as a processing apparatus) 1 in the first embodiment.
- the processing device 1 is a so-called computer and includes, for example, a CPU (Central Processing Unit) 2, a memory 3, a communication unit 4, an input / output interface (I / F) 5 and the like that are connected to each other via a bus.
- the memory 3 is a RAM (Random Access Memory), a ROM (Read Only Memory), a hard disk, or the like.
- the communication unit 4 exchanges signals with other computers and devices.
- a monitoring camera 7 is connected to the communication unit 4.
- the communication unit 4 acquires a video signal from the monitoring camera 7.
- a portable recording medium or the like can be connected to the communication unit 4.
- the surveillance camera 7 is installed at a position and orientation where an arbitrary place to be monitored can be photographed, and sends the photographed video signal to the communication unit 4.
- a place taken by the monitoring camera 7 may be referred to as a monitoring target place.
- FIG. 1 one monitoring camera 7 is shown, but a plurality of monitoring cameras may be connected to the communication unit 4.
- the input / output I / F 5 can be connected to a user interface device such as the output device 6.
- the output device 6 is any one or a plurality of display devices, printing devices, projection devices, and the like.
- the display device is a device that displays a screen corresponding to drawing data processed by a CPU 2 or a GPU (Graphics Processing Unit) (not shown) such as an LCD (Liquid Crystal Display) or CRT (Cathode Ray Tube) display. It is.
- the input / output I / F 5 may be connected to an input device (not shown) that receives input of a user operation such as a keyboard and a mouse. Further, a touch panel in which the output device 6 and the input device are integrated may be connected to the input / output I / F 5.
- the hardware configuration of the processing device 1 is not limited.
- FIG. 2 is a diagram conceptually illustrating a processing configuration example of the image processing apparatus 1 in the first embodiment.
- the processing device 1 includes an acquisition unit 11, a person detection unit 12, a change detection unit 13, a holding unit 14, a determination unit 15, an output processing unit 16, and the like.
- Each of these processing units is realized, for example, by executing a program stored in the memory 3 by the CPU 2.
- the program is installed from a portable recording medium such as a CD (Compact Disc) or a memory card or another computer on the network via the communication unit 4 or the input / output I / F 5 and is stored in the memory 3. It may be stored.
- a portable recording medium such as a CD (Compact Disc) or a memory card or another computer on the network via the communication unit 4 or the input / output I / F 5 and is stored in the memory 3. It may be stored.
- CD Compact Disc
- the acquisition unit 11 acquires an input image. Specifically, the acquisition unit 11 sequentially acquires input images by capturing a video signal from the monitoring camera 7 at an arbitrary timing.
- the arbitrary timing is, for example, a predetermined cycle.
- the input image acquired by the acquisition unit 11 is held in the holding unit 14.
- the person detection unit 12 detects a person from the input image acquired by the acquisition unit 11.
- the person detection unit 12 may detect the whole body of the person, or may detect a part of the person such as the head, face, and upper body.
- a person detection range by the person detection unit 12 depends on detection contents of a state change by a change detection unit 13 described later. This dependency will be described later.
- the person detection unit 12 detects a person using a known image recognition technique.
- the person detection unit 12 holds an image feature amount corresponding to a human detection range, and detects a region similar to the feature amount in the input image as the detection range.
- the detection method by the person detection unit 12 is not limited.
- the change detection unit 13 detects a state change of the target in the person detected by the person detection unit 12 based on a plurality of input images sequentially acquired by the acquisition unit 11. For example, the change detection unit 13 detects a change in the appearance of the person to the target state while tracking (tracking) the person detected by the person detection unit 12 between a plurality of input images. As a method for tracking a person between a plurality of images, a well-known tracking method for an object and a person may be used. For example, the change detection unit 13 compares the feature amounts of the respective human areas detected in the respective images, and recognizes the approximate human areas as the same person. The person tracking method by the change detection unit 13 is not limited.
- the change detection unit 13 can hold a state after the change in the state change of the target in advance, and can detect the state change of the target using the held state after the change as follows. .
- the state after the change is held in advance by an image feature amount, for example.
- the change detection unit 13 determines whether or not the state of the person detected by the person detection unit 12 is in the held state in the input image.
- the change detection unit 13 uses the input image captured before the input image to determine whether the state of the person has changed. Judging.
- the change detection unit 13 detects any state change in tracking a person between input images, and when a change is detected, determines whether or not the state after the change is in a state to be held in advance. May be.
- the state change detection method by the change detection unit 13 is not limited to such a method.
- the state change of the target can be detected without performing tracking between a plurality of images.
- the change detection unit 13 may detect a state where the foot is floating from the ground or the floor in one input image.
- the state change is detected from one input image. Can do.
- a change in the state of a person in the input image can occur due to various factors. For example, even if a person is walking, his / her legs move, and the posture (state) of the person changes between images. In addition, when a person watches a smartphone while walking, a posture change due to walking and a posture change of watching the smartphone occur.
- the change detection unit 13 detects a change in the appearance of the person caused by the occurrence of the abnormal state among all the state changes as the state change of the target.
- the state change of the target may include a change in posture, a change in movement, an appearance change expressed by emotions, and the like.
- the state change of the subject is squatting, turning around, looking up, holding the head, pointing the finger, shooting, running out, falling down, falling down, jumping, standing up, frowning, staring, surprised face There may be.
- the state change of the object is a change in human appearance, and the specific content of the change is not limited as long as the change is caused by the occurrence of an abnormal state.
- the state change that can be detected may be limited depending on the installation position of the monitoring camera 7 that captures the input image, the number of pixels of the monitoring camera 7 (image resolution), and the like.
- a change in facial expression or a part of the face appears as a change in a small image area, and may not be detected.
- the change detection unit 13 may detect a change in the appearance of the rough part of the body, such as the head, the trunk, the upper limbs, and the lower limbs, or the entire body as the state change of the target. Even if the change in the appearance of the whole body or a rough part of the body is the change in the state of the target, it is possible to detect squatting, falling down, running out, pointing, etc., so the abnormal state can be determined. is there.
- the person detection range by the person detection unit 12 is determined in advance according to the state change of the target to be detected by the change detection unit 13. For example, in order to detect a state change such as turning around, looking up, or holding the head, the person detection unit 12 may detect the head or upper body. In order to detect a change in state such as frowning, closing eyes, blushing, or pale, the person detection unit 12 may detect the face. In order to detect a state change such as squatting, falling, or jumping, it is desirable that the person detection unit 12 detects the entire image of the person.
- state change is illustrated.
- a person's abnormal state occurs, such as a person's suspicious state
- the people around the person turn around, point at their fingers, run out, grimac, or make a surprised face.
- a person around the object is likely to cause a state change such as jumping, turning around, frowning, or pinching the nose.
- an abnormal situation in the field such as a fire occurs
- the people around the scene will start running, operate the mobile phone, point to the finger, take a picture, turn around, and shake hands.
- the holding unit 14 holds each input image acquired by the acquisition unit 11, and further holds information (state change information) for each state change of the target detected by the change detection unit 13.
- FIG. 3 is a diagram illustrating an example of the state change information held in the holding unit 14.
- the state change information includes information such as an occurrence time, a person ID, and contents of the state change.
- the generation time the acquisition time of the input image in which the changed state is detected or the identification number of the input image is set.
- the person ID is identification data given to a person every time a new person is detected by the person detection unit 12.
- FIG. 4 is a diagram illustrating another example of the state change information held in the holding unit 14.
- the position of the person corresponding to the person ID on the image may be set in the state change information.
- Such a position is specified by, for example, the person detection unit 12 and set in the state change information held in the holding unit 14.
- This position is determined as a representative position of the person detected by the person detection unit 12, such as the center position of the head and the center position of the chest.
- camera parameters indicating the position, posture, focal length, etc. of the monitoring camera 7 are held, the three-dimensional position of the person (real space position) using the camera parameters from the position of the person on the image shown above. And the position may be added and set. Further, when a person's position is handled in the processing described later, this person's three-dimensional position may be used.
- the determination unit 15 detects that the target state change has occurred by a plurality of people, and determines an abnormal state according to this detection. Specifically, the determination unit 15 constantly monitors the state change information held in the holding unit 14 and detects that the target state change has occurred in a predetermined number or more based on the state change information.
- the predetermined number is held in advance by the determination unit 15.
- the predetermined number of persons may be two or more, and is determined according to the number of persons who can be reflected in the acquired input image. However, the accuracy of determining an abnormal condition increases as the predetermined number of people increases. The fact that the state change of the object as described above has occurred in many people is because there is a high possibility that some abnormal state has occurred.
- the determination unit 15 determines that some abnormal state exists or has occurred when it is detected that the target state change has occurred by a plurality of persons. There may be one or more types of target state changes that are the basis for the abnormal state determination. For example, the determination unit 15 sums up the number of persons who have occurred for each of a plurality of types of target state changes such as squatting, falling down, holding their heads, and an abnormal condition exists if the total number of persons is greater than or equal to a predetermined number. It may be determined that it has occurred or has occurred. In addition, the determination unit 15 may determine that an abnormal state exists or has occurred when there are more than a predetermined number of persons counted for each state change of the target.
- the determination unit 15 can further use a shift in the occurrence timing of the state change of the target among a plurality of persons for determining an abnormal state. In this case, when the determination unit 15 detects that the target state change has occurred in a predetermined number or more, the determination unit 15 calculates the time width of the set of occurrence times of the state change. The determination unit 15 determines whether the time width is within a predetermined time width. That is, the determination unit 15 detects that a change in the state of the target has occurred within a predetermined time by a plurality of persons, and determines an abnormal state according to the detection result.
- the determination accuracy of the abnormal state can be improved.
- the predetermined time is held in advance by the determination unit 15.
- the predetermined time can also be expressed using the number of input images.
- the number of input images can be converted into time according to the period of obtaining the input image from the video signal sent from the monitoring camera 7.
- 10 input images correspond to 5 seconds. Therefore, when the state change information held in the holding unit 14 includes the identification number of the input image as the occurrence time as described above, the predetermined number of input images is used as the predetermined time.
- the determination unit 15 can also estimate the degree of the abnormal state according to the difference in the occurrence timing of the state change of the target among a plurality of persons.
- the degree of the abnormal state to be estimated is the magnitude of the abnormal state, the high degree of abnormality, the high degree of risk associated with the abnormal state, and the like. If the deviation is small, the determination unit 15 can determine that the degree of the abnormal state is large. When the degree of the abnormal state is large, the reaction speed of the surrounding person to the abnormal state is increased. For example, when a dangerous person enters the monitoring target location, a person who notices the dangerous person immediately reacts greatly, and the reaction is rapidly propagated to surrounding persons depending on the magnitude of the reaction.
- the determination unit 15 can estimate that the degree of the abnormal state is small and the abnormal state continues to some extent.
- the determination unit 15 can detect that a plurality of persons have changed the state of the target with the same state after the change, and can also determine an abnormal state according to the detection. In this case, the determination unit 15 counts the number of persons with the same state change content in the state change information held in the holding unit 14, and there is a state change in which the number of persons is equal to or greater than a predetermined number. It is determined that there is an abnormal condition. When multiple people have caused the same state change, there is a high possibility that some abnormal state has occurred. Therefore, by further adding the condition “the state after change is the same” to the abnormal state determination, the determination accuracy of the abnormal state can be improved.
- the determination unit 15 can detect that a change in the state of the target has occurred in a plurality of persons close in distance, and can determine an abnormal state in accordance with this detection. In this case, the determination unit 15 uses the position information of the state change information held in the holding unit 14 to calculate the relative distance (intra-image distance) of each person who caused the target state change. The determination unit 15 counts the number of persons whose calculated relative distance is within a predetermined distance, and determines that there is an abnormal condition when the number of persons is equal to or greater than the predetermined number. When a plurality of persons close to each other cause a change in the state of the target, there is a high possibility that some abnormal state has occurred. Therefore, by further adding the condition “the relative distance of each person who caused the target state change is short” to the abnormal state determination, the determination accuracy of the abnormal state can be improved.
- each method may be appropriately selected according to the type of abnormal state to be detected. For example, when a plurality of types of abnormal states are desired to be detected as much as possible, the determination unit 15 determines that there is an abnormal state because the state change of the target has occurred in a predetermined number or more. When it is desired to further increase the determination accuracy, the determination unit 15 may execute each of ⁇ other determination methods 1, 2, and 3 of the abnormal state >> and determine the abnormal state according to each determination result. . Further, in order to further improve the determination accuracy, the determination unit 15 is close in distance to a state change of an object having the same state after a change within a predetermined time when all the above-described conditions are satisfied. It is only necessary to determine that there is an abnormal condition only when more than a predetermined number of people wake up. However, the conditions used in each method may be combined as appropriate.
- the output processing unit 16 can cause the output device 6 to output the presence of an abnormal state indicated by the determination result by the determination unit 15. For example, the output processing unit 16 sends the video signal from the monitoring camera 7 to the output device 6 and causes the output device 6 to always display the video imaged by the monitoring camera 7. When the determination unit 15 determines that there is an abnormal state, the output processing unit 16 superimposes and displays a drawing element indicating the presence of the abnormal state on the display image. The output processing unit 16 can also arrange the images indicating that there is an abnormal state and output an image before the abnormal state is determined. Further, the output processing unit 16 can output the presence of an abnormal state by using an output device (LED (Light Emitting Diode) or the like) different from the output device 6 that displays the display image.
- LED Light Emitting Diode
- the presence of an abnormal state can be output by turning on the LED, changing the lighting color of the LED, or the like.
- the output form is not limited to display or lighting, and may be, for example, printing, data file generation, or the like, and the output form is not limited.
- the output processing unit 16 can cause the output device 6 to output the following contents in addition to the presence or absence of an abnormal state or instead of the presence or absence of an abnormal state.
- drawing elements including number information, order information, coloring, etc.
- drawing elements may be added by editing the input image held in the holding unit 14, or may be added to the video signal from the monitoring camera 7. It may be attached in a form superimposed on the video based on it.
- the output processing unit 16 causes the output device 6 to output an image to which a drawing element is attached to a person group in which detection of a state change in the input image is based on the abnormal state determination or each person in the person group.
- the position to which the drawing element is attached is determined by the position information of the state change information held in the holding unit 14, for example.
- a group of people whose detection of a change in state is a basis for determining an abnormal state may be referred to as a “group of people as a basis”.
- the output processing unit 16 causes the output device 6 to output an image obtained by attaching the drawing element to the input image held in the holding unit 14.
- the drawing element is, for example, a line surrounding a base person group, some mark (point, star, exclamation mark, arrow, balloon, etc.) attached to the person group or each person of the person group.
- the drawing element itself, the position to which the drawing element is attached, and the way of attaching the drawing element are not limited as long as the group of persons as a basis or each person can be specified.
- FIG. 5 is a diagram illustrating a first example of an output image of the output device 6.
- FIG. 5 shows a situation in which a dangerous person D1 appears at a monitoring target location photographed by the monitoring camera 7, and the processing apparatus 1 detects an abnormal state of the person (also considered an abnormal state of the field).
- the processing apparatus 1 detects an abnormal state of the person (also considered an abnormal state of the field).
- the processing apparatus 1 detects an abnormal state of the person (also considered an abnormal state of the field).
- the processing apparatus 1 detects an abnormal state of the person (also considered an abnormal state of the field).
- the processing apparatus 1 detects an abnormal state of the person (also considered an abnormal state of the field).
- the processing apparatus 1 detects an abnormal state of the person (also considered an abnormal state of the field).
- the processing apparatus 1 detects an abnormal state of the person (also considered an abnormal state of the field).
- the processing apparatus 1 detects an abnormal state of the person (also considered an abnormal state of the field).
- the processing apparatus 1 detects an abnormal state
- passers-by M3 to M13 notice the dangerous person D1, causing a change in the state of the target. Specifically, passers-by M1 to M8 have changed from a walking state to a state of running, passers-by M9 to M11 have changed from a walking state to a state of turning, and passers-by M12 and M13 are in a walking state. It has changed to a state where it stops. Other passersby are still walking normally.
- the determination unit 15 detects from the input image 13 persons who have started to change, turn around, and stop, and determine that there is an abnormal state according to this detection. That is, in the example of FIG.
- the determination unit 15 holds the predetermined number “3”, and determines that there is an abnormal state when the target state change occurs in three or more people.
- the output processing unit 16 causes the output device 6 to output an image with a rectangular drawing element P1 surrounding them to the person groups M1 to M13 that are the basis of the abnormal state determination.
- an abnormal state (dangerous person D1) exists in the drawing element P1
- a person who looks at the output image can immediately confirm the position of the abnormal state and its contents.
- FIG. 6 is a diagram showing a second example of the output image of the output device 6.
- the determination unit 15 determines the abnormal state using the above-mentioned ⁇ other determination method 2 of abnormal state >>. Specifically, the determination unit 15 counts the number of persons who have caused the state change of the target having the same state after the change. The number of people who caused the state change “run out” is counted as “8” (passers M1 to M8), and the number of people who caused the subject state change “turn around” is “3” (passers M9 to M11). And the number of persons who have caused the target state change “stop” is counted as “2 persons” (passengers M12 and M13).
- the determination unit 15 determines that there is an abnormal state because the target state changes “run out” and “turn around” occur with a predetermined number of people “3” or more.
- the output processing unit 16 outputs an image to which a rectangular drawing element is attached for each person group having the same state after the change.
- the drawing element P1 surrounds a group of people whose changed state is “run out”
- the drawing element P2 surrounds a group of persons whose changed state is “turn around”
- the drawing element P3 has a changed state. Enclose a group of people who will “stop”.
- a portion where each drawing element overlaps is indicated by a dotted line.
- the output processing unit 16 selects either the number of persons in the person group and the relative distance of each person in the person group with respect to the person group whose detection of the state change is the basis of the abnormal state determination in the input image, or It is also possible to cause the output device 6 to output an image with both added.
- the number of persons and the relative distance are given to positions where the target person group can be specified.
- the number of persons in the basis of the person group and the relative distance of each person in the person group represent the determination accuracy of the abnormal state. This is because there is a high possibility that an abnormal condition exists when the number of persons who have caused a change in the state of the object is large and when the distance between the persons who have caused the change in the state of the object is short. Therefore, such an output image enables a person who has seen the image to easily estimate the determination accuracy of the abnormal state.
- the number of persons in the basis of the group of people may further indicate the degree of abnormal condition.
- the degree of abnormal state may be large. Therefore, according to the output image to which the number of people is attached, the person who viewed the image can also estimate the degree of the abnormal state.
- FIG. 7 is a diagram illustrating a third example of the output image of the output device 6.
- the output processing unit 16 causes the output device 6 to output an image in which the number of people “13” is added together with a rectangular drawing element to the basis person group.
- the output processing unit 16 is an image in which the state change order or the time change of the state change is attached to each person of the group of persons whose detection of the state change is the basis of the abnormal state determination as the drawing element or together with the drawing element. Can also be output to the output device 6.
- the time of the state change of each person the time corresponding to each input image in which the state change is detected is used.
- the state change time difference is a time difference between the time when the earliest state change occurs and the state change time caused by each person. For example, “ ⁇ 0” is added around the person who caused the first state change, and the time difference between the other person around the time of the state change caused by that person “+5 seconds”, “+8 seconds” Is attached.
- FIG. 8 is a diagram showing a fourth example of the output image of the output device 6.
- the output processing unit 16 outputs an image in which a rectangular drawing element is attached to a base person group and each person in the person group is assigned a state change order. 6 to output.
- a different order is assigned to each person, but the same order may be assigned to each person whose target state change is detected in the same input image.
- the output processing unit 16 can also cause the output device 6 to output an image to which a drawing element indicating a direction corresponding to the changed state in the state change of a plurality of persons, which is the basis for the abnormal state determination, is attached.
- the drawing element representing this direction is attached to the grounded person group or each person of the person group.
- the output processing unit 16 holds in advance an association between the changed state and the corresponding direction information. For example, when the state after the change is “squatting”, the abnormal state is likely to exist above or below the person, so “upward” and “downward” with respect to the changed state “squatting”. "Is associated.
- the output processing unit 16 specifies direction information corresponding to the state after the change using such association information, and selects a drawing element corresponding to the specified direction information. Examples of the drawing element indicating the direction include an arrow.
- FIG. 9 is a diagram showing a fifth example of the output image of the output device 6.
- the output processing unit 16 adds a rectangular drawing element to the ground person group, and the drawing element indicating the direction corresponding to the changed state for each person of the person group.
- Each of the attached images is output to the output device 6.
- the passersby M1, M6, M7, and M8 are given an arrow (the left direction in the drawing) opposite to the “start” direction (the right direction in the drawing), and the passersby M2 to M5 have the “start” direction (the left in the drawing).
- An arrow (right direction in the drawing) opposite to the direction is attached.
- passers-by M10 and M11 are provided with an arrow in the left direction as the “turn around” direction
- the passer-by M9 is provided with an arrow in the right direction in the drawing as the “turn around” direction.
- passers-by M12 and M13 are provided with arrows in the left and right directions in the drawing as directions in which the face faces in the “stop” state.
- Output example 5 The output processing unit 16 outputs an image in which the input image is colored in a color corresponding to the distance from the representative position in a circle or an ellipse centered on the representative position on the input image of the group of persons as a basis. It can also be output to the device 6.
- the processing apparatus 1 may have the following processing configuration.
- FIG. 10 is a diagram conceptually illustrating a processing configuration example of the image processing apparatus 1 in the modification of the first embodiment.
- the processing device 1 further includes a determination unit 17 in addition to the processing configuration illustrated in FIG. 2.
- the determination unit 17 is also realized in the same manner as other processing units.
- the determination unit 17 determines a representative position on the input image of the group of persons whose detection of the state change is the basis for the abnormal state determination.
- the determination unit 17 uses the position information of the state change information held in the holding unit 14 to specify the position of each person in the basis person group, and based on the specified position, the center of gravity position of the person group Can be determined as the representative position.
- the representative position only needs to indicate a representative position of the person group as a basis, and is not limited to the position of the center of gravity.
- the output processing unit 16 may continue the output of the image for a duration corresponding to the number of persons in the group of persons whose detection of the state change is the basis for the abnormal state determination. As described above, the more people who have caused the target state change, the higher the possibility that an abnormal state actually exists. Therefore, the larger the number of persons in the base person group, the longer the image is output, thereby making it easier for the person viewing the image to notice the abnormal state.
- FIG. 11 is a flowchart showing an operation example of the processing apparatus 1 in the first embodiment.
- the abnormality detection method in the first embodiment is executed by at least one computer such as the processing device 1.
- each illustrated process is executed by each processing unit included in the processing apparatus 1. Since each process is the same as the processing content of the above-mentioned each process part which the processing apparatus 1 has, the detail of each process is abbreviate
- the processing apparatus 1 acquires an input image (S91).
- the input image is sequentially acquired by capturing the video signal from the monitoring camera 7 at an arbitrary timing.
- the processing apparatus 1 detects a person from the acquired input image (S92).
- the processing apparatus 1 may detect the whole body of a person or a part of the person.
- the processing apparatus 1 detects a change in the state of the target in the person detected in (S92) based on the input image acquired in (S91) (S93).
- the change in the state of the subject means a change in the appearance of the person caused by the occurrence of an abnormal state.
- the processing apparatus 1 holds information (state change information) for each state change of the target detected in (S93) (S94).
- the state change information includes information such as the occurrence time, the person ID, and the contents of the state change. Further, the state change information may include a position on the person's image indicated by the person ID in addition to them. Further, the state change information may include a position of a person in real space.
- the processing apparatus 1 refers to the state change information held to determine whether or not the target state change has occurred in a predetermined number or more (S95).
- the processing device 1 determines that there is an abnormal state (S96).
- the processing apparatus 1 can further use the deviation of the occurrence timing of the state change of the target with a predetermined number or more for the determination of the abnormal state.
- the processing device 1 determines that there is an abnormal state when a change in the state of the target occurs in a predetermined time by a predetermined number or more. By using this generation time difference, the processing apparatus 1 can also estimate the degree of the abnormal state.
- the processing device 1 causes the output device 6 to output the determination result of the abnormal state (S97).
- the processing device 1 may cause the output device 6 to output the presence / absence of an abnormal state, or together with the presence / absence of the abnormal state, or instead of the presence / absence of the abnormal state, the image shown in the above ⁇ Output Examples 1 to 5 >>. Can also be output.
- FIG. 11 a plurality of steps (processes) are shown in order, but the execution order of each step in the abnormality detection method in the first embodiment is not limited to the example of FIG.
- (S94) is executed only when both (S92) and (S93) have been successfully detected.
- (S91) can be executed at a predetermined cycle regardless of the determination result of (S95).
- a further condition may be added as shown in ⁇ Other determination methods 1 to 3 of abnormal state >> described above.
- the processing device 1 determines whether or not a target state change has occurred in a predetermined number of people or more within a predetermined time.
- processing device 1 determines whether or not the state change of the same target after the change has occurred in a predetermined number or more. To do.
- ⁇ another determination method 3 of abnormal state >> is executed, in (S95), the processing device 1 determines whether or not the target state change has occurred in a predetermined number of persons or more close in distance. Further, the processing apparatus 1 may determine a condition in which any one of these conditions is combined in (S95).
- the first embodiment when a change in the state of a target in a person is detected based on the input image and the change in the state of the target occurs in a predetermined number or more, it is determined that there is an abnormal state. If an abnormal state of a person such as a suspicious state of a person, an abnormal state of an object such as an abandoned object or filth, or an abnormal state of a field such as a fire or an accident occurs, there is a person around the location When doing so, the people change appearances such as postures, behaviors, facial expressions, etc. that are not normally raised at the monitored location. Therefore, in the first embodiment, various abnormal states in a person, an object, a field, and the like are captured by changes in the state of the people around them.
- various abnormal states such as a person, an object, and a field can be comprehensively detected. Furthermore, according to the first embodiment, if a change in the state of surrounding people can be captured, the abnormal state can be detected even if the abnormal state itself is not reflected in the input image. For example, even when an abnormal state exists outside the imaging range of the monitoring camera 7 or when the abnormal state is hidden by an obstacle, the abnormal state can be detected according to the first embodiment. Furthermore, according to the first embodiment, in the output image, by adding some drawing element to the person group that is the basis of the abnormal state determination or each person of the person group, the presence of the abnormal state is given to the person who views the image. The position can be estimated.
- the processing apparatus 1 in the second embodiment controls the monitoring camera 7 using the determination result of the abnormal state.
- the processing apparatus 1 in the second embodiment will be described focusing on the content different from the first embodiment. In the following description, the same contents as those in the first embodiment are omitted as appropriate.
- FIG. 12 is a diagram conceptually illustrating a processing configuration example of the image processing apparatus 1 in the second embodiment.
- the processing apparatus 1 further includes a control unit 18 in addition to the processing configuration in the first embodiment.
- the control unit 18 is realized in the same manner as other processing units.
- the control unit 18 controls the monitoring camera 7 using the state change information held in the holding unit 14.
- a specific example of the control method of the monitoring camera 7 by the control unit 18 will be described.
- the control unit 18 can control the imaging direction of the monitoring camera 7 based on the representative position on the input image of the group of persons whose detection of the state change is the basis for the abnormal state determination.
- the processing device 1 further includes a determination unit 17. As described above, the determination unit 17 determines the representative position on the input image of the group of persons whose detection of the state change is the basis for the abnormal state determination.
- the surveillance camera 7 includes a mechanism (not shown) for adjusting the imaging direction.
- This mechanism may change the direction of the monitoring camera 7, or may change the imaging direction by a movable mirror or an optical system.
- the control unit 18 instructs the monitoring camera 7 or the mechanism to change the imaging direction via the communication unit 4. Specifically, the control unit 18 instructs the change of the imaging direction so that the representative position of the person group as the basis is in the center of the input image. In response to this instruction, the imaging direction of the surveillance camera 7 is adjusted.
- the representative position of the person group that is used as the basis for the abnormal state determination is located at the center of the input image, so that it is easy for the viewer who sees the image to check the representative position of the person group that serves as the basis and the surrounding area. Become. Thereby, since there is a high possibility that an abnormal state exists in the position of the person group serving as a basis and in the vicinity thereof, the location of the abnormal state can be easily estimated.
- the control unit 18 can control the angle of view of the monitoring camera 7 based on the size on the input image of the group of persons whose detection of the state change is the basis for the abnormal state determination. For example, the control unit 18 monitors an instruction to change the angle of view of the monitoring camera 7 so that the basis group of persons falls within a predetermined range in a part of the input image and the periphery of the group of persons is reflected in the input image. Send to camera 7. In response to this instruction, the surveillance camera 7 zooms in or out. The size of the person group on the input image can be determined, for example, by adding together the image areas of the persons detected by the person detection unit 12. The control unit 18 may continue to send an instruction to zoom in or zoom out to the monitoring camera 7 until the size of the person group falls within the predetermined range.
- the control unit 18 can continuously change the imaging direction of the monitoring camera 7 within a range in which the whole or a part of the group of persons whose detection of the state change is the basis of the abnormal state determination is included in the input image. Specifically, the control unit 18 recognizes the position of the basis person group on the input image based on the state change information held in the holding unit 14. Based on the recognized position, the control unit 18 can calculate an imaging direction change range corresponding to a range in which at least a part of the person group fits in the input image. The control unit 18 continuously sends an instruction to change the imaging direction to the monitoring camera 7 or a mechanism for adjusting the imaging direction so that the position of the person group on the input image changes within the change range. For example, the control unit 18 includes the direction to be changed (up and down, left and right, etc.) and the angle to be changed in the imaging direction change instruction.
- This method makes it possible to comprehensively check the situation around the person group that is the basis.
- an abnormal state that is not reflected in the input image when it is determined that there is an abnormal state can be included in the shooting range of the monitoring camera 7.
- the control unit 18 can also change the imaging direction of the monitoring camera 7 in the direction corresponding to the changed state in the state change of a plurality of persons, which is the basis for the abnormal state determination.
- the direction corresponding to the state after the change and the method for acquiring the direction information are as described in ⁇ Output Example 4 >>.
- the control unit 18 acquires direction information corresponding to the changed state, and sends an imaging direction change instruction including the direction information to the monitoring camera 7 or a mechanism that adjusts the imaging direction.
- the imaging direction of the monitoring camera 7 can be directed to the direction in which the abnormal state is inferred from the state after the change in the state change of the target. Can be confirmed.
- an abnormal state that is not reflected in the input image when it is determined that there is an abnormal state can be included in the shooting range of the monitoring camera 7.
- the output processing unit 16 may continue the control of the monitoring camera 7 for a duration corresponding to the number of persons in the group of persons whose detection of the state change is the basis for the abnormal state determination.
- the control of the monitoring camera 7 as described above is continued longer as the number of person groups serving as a basis increases.
- the monitoring camera 7 is controlled so that the imaging direction changed based on the representative position of the person group as a basis and the changed state is maintained for a long time.
- the monitoring camera 7 is controlled so that the angle of view changed based on the size of the basis person group on the input image is maintained long.
- the continuous change in the imaging direction of the monitoring camera 7 based on the position of the person group serving as the basis continues for a long time. This makes it easier for a person viewing the image to notice the abnormal state.
- FIG. 13 is a flowchart showing an operation example of the processing apparatus 1 in the second embodiment.
- steps having the same processing contents as those in FIG. 11 are denoted by the same reference numerals as those in FIG.
- the execution subject of the abnormality detection method in the second embodiment is the same as in the first embodiment.
- (S131) is further executed with respect to the method in the first embodiment.
- the processing device 1 controls the monitoring camera 7 using the state change information held in the holding unit 14.
- the specific control method is as described above.
- the imaging direction of the monitoring camera 7 is based on the position of the person group as a basis, the state after each person of the person group is changed, and the like. And the angle of view is controlled. As a result, it is possible to visually recognize the situation around the person group that is the basis of the possibility that an abnormal state exists, so that the abnormal state can be easily confirmed.
- the processing device 1 is connected to a plurality of monitoring cameras 7 and can also control the plurality of monitoring cameras 7.
- each monitoring camera 7 is installed such that different monitoring target locations are in the shooting range.
- the processing device 1 corresponds to the position of the group of persons that is the basis for the abnormal state determination. May be selected and the imaging direction and angle of view of the monitoring camera 7 may be changed. In this way, even if the imaging direction and the angle of view of one monitoring camera 7 are changed as described above, even if an abnormal state exists at a position where the monitoring camera 7 cannot display the image, By controlling the monitoring camera 7, the position can be imaged.
- a plurality of processing devices 1 respectively connected to at least one monitoring camera 7 may cooperate with each other.
- the processing device 1 detects an abnormal state
- the processing device 1 provides an abnormal state determination result to other processing devices 1 in the vicinity.
- the processing device 1 changes the imaging direction and angle of view of its own monitoring camera 7 so as to photograph the surroundings of the monitoring target location of the monitoring camera 7 of the processing device 1 of the providing source.
- the processing device 1 can also provide information regarding the position of the person group serving as the basis to other processing devices 1.
- the monitoring camera 7 may be used in which the processing device 1 is not connected to the processing device 1 and the processing device 1 is connected to another computer.
- the acquisition unit 11 may acquire an input image or a video signal from a portable recording medium or another computer via the communication unit 4.
- the other computer include an image storage / distribution device that temporarily stores and distributes images captured by a camera, and an image recorder that stores and reproduces images.
- the output processing unit 16 may output the image to another computer via the communication unit 4.
- the control unit 18 transmits instruction information for controlling the monitoring camera 7 to the other computer.
- FIG. 14 is a diagram conceptually illustrating a processing configuration example of the image processing apparatus according to the third embodiment.
- the image processing apparatus 100 includes a change detection unit 101 and a determination unit 102.
- the image processing apparatus 100 has, for example, the same hardware configuration as the above-described processing apparatus 1 illustrated in FIG. 2, and the above-described processing units are realized by processing a program in the same manner as the processing apparatus 1. Is done. However, the image processing apparatus 100 may not be connected to the output device 6 and the monitoring camera 7.
- the change detecting unit 101 detects a change in the state of the target in the person based on the input image.
- the specific processing content of the change detection unit 101 is the same as that of the change detection unit 13 described above.
- the change detection unit 101 acquires information about an image area of a person on each input image from the acquisition unit 11 and the person detection unit 12 realized by another computer, and uses this information. Thus, it is also possible to detect a change in the state of an object in a person.
- the image processing apparatus 100 according to the third embodiment does not necessarily require the acquisition unit 11 and the person detection unit 12. Further, the change detection unit 101 can cause another computer to hold information regarding the detected state change of the target.
- the determination unit 102 detects that the target state change has occurred by a plurality of persons, and determines an abnormal state in response to this detection.
- the specific processing content of the determination unit 102 is the same as that of the determination unit 15 described above.
- the determination unit 102 can acquire the state change information from the holding unit 14 on another computer and use it for abnormal state determination. Further, the determination unit 102 can notify the determination result to another computer.
- FIG. 15 is a flowchart illustrating an operation example of the image processing apparatus 100 according to the third embodiment.
- the abnormality detection method in the third embodiment is executed by at least one computer such as the image processing apparatus 100.
- each illustrated process is executed by each processing unit included in the image processing apparatus 100.
- the abnormality detection method in the third embodiment detects a change in the state of a target in a person based on an input image (S151), and detects that a change in the state of the target has occurred in a plurality of people (S152; YES).
- An abnormal state is determined (S153).
- (S151) corresponds to (S93) in FIGS. 11 and 13
- (S152) corresponds to (S95) in FIGS. 11 and 13
- (S153) corresponds to (S96) in FIGS. It corresponds to.
- the third embodiment may be a program that causes at least one computer to execute such an abnormality detection method, or a recording medium that can be read by at least one computer that records such a program. May be.
- Change detecting means for detecting a change in the state of the object in the person based on the input image; Detecting means for detecting that the state change of the target has occurred in a plurality of persons, and determining an abnormal state in response to the detection;
- An image processing apparatus comprising: 2. The determination means detects that the state change of the target having the same state after the change has occurred in a plurality of people. 1. An image processing apparatus according to 1. 3. The determination means detects that the state change of the object has occurred in a plurality of persons close in distance; 1. Or 2. An image processing apparatus according to 1. 4). The determination means further uses a shift in the occurrence timing of the state change of the target among the plurality of persons for determining an abnormal state, 1. To 3.
- the image processing apparatus according to any one of the above. 5.
- the determination means estimates the degree of an abnormal state according to a shift in the occurrence timing of the state change of the target in the plurality of persons. 4).
- An image processing apparatus according to 1. 6).
- 1. Output processing means for causing an output device to output an image to which a drawing element has been attached to a group of persons whose detection of a state change in an input image is the basis for determining an abnormal state.
- the image processing apparatus according to any one of the above. 7).
- the output processing means for the person group whose detection of the state change is the basis of the abnormal state determination in the input image, either the number of persons in the person group and the relative distance of each person in the person group or Causing the output device to output the image to which both have been added, 6).
- An image processing apparatus according to 1. 8).
- the output processing means includes, as the drawing element or together with the drawing element, each person in the group of persons whose detection of the state change is the basis for the abnormal state determination, the order of the state change or the time difference of the state change. Outputting an image to the output device; 6). Or 7.
- the output processing means causes the output device to output the image to which the drawing element indicating the direction corresponding to the changed state in the state change of a plurality of persons, which is the basis of the abnormal state determination, is output. 6).
- Determining means for determining a representative position on the input image of the group of persons whose detection of the state change is the basis for the abnormal state determination;
- Output processing means for causing the output device to output an image in which the input image is colored in a color corresponding to the distance from the representative position in a circle or ellipse centered on the representative position; 1 is further provided.
- the image processing apparatus according to any one of the above. 11.
- the output processing means continues the output of the image for a duration corresponding to the number of persons in the group of persons whose detection of the state change is the basis for the abnormal state determination. 6).
- the image processing apparatus according to any one of the above. 12 Determining means for determining a representative position on the input image of the group of persons whose detection of the state change is the basis for the abnormal state determination; Control means for controlling the imaging direction of the imaging device based on the representative position; 1 is further provided.
- Control means for controlling the angle of view of the imaging device based on the size on the input image of the group of persons whose detection of the state change is the basis for the abnormal state determination; 1 is further provided.
- an abnormality detection method executed by at least one computer, Detects a change in the target state of a person based on the input image, Detecting that the subject's state change has occurred in multiple persons; Determining an abnormal state in response to the detection; Anomaly detection method including 18.
- the detection that the state change of the target has occurred in a plurality of people is to detect that the state change of the target having the same state after the change has occurred in a plurality of people, 17.
- the abnormality detection method described in 1. 19. Detecting that the state change of the object has occurred in a plurality of people, detecting that the state change of the object has occurred in a plurality of people close in distance; 17. Or 18.
- the detection that the state change of the object has occurred in a plurality of persons further determines that a shift in the occurrence timing of the state change of the object in the plurality of persons is within a predetermined time. 17.
- the abnormality detection method as described in any one of these. 21.
- Inferring the degree of abnormal state according to the deviation of the occurrence time of the state change of the target in the plurality of people Further includes: The abnormality detection method described in 1. 22. 18. further including causing the output device to output an image with a drawing element to a group of persons whose detection of a state change in the input image is the basis for determining an abnormal state or each person of the group of persons.
- the abnormality detection method as described in any one of these. 23.
- the attached image is output to the output device. 22.
- the output includes, as the drawing element or together with the drawing element, each person in the group of persons whose detection of the state change is the basis of the abnormal state determination, the image in which the order of the state change or the time difference of the state change is attached. Outputting to the output device; 22. Or 23.
- the output causes the output device to output the image to which the drawing element indicating the direction corresponding to the state after the change in the state change of a plurality of persons that is the basis of the abnormal state determination is provided. 22. To 24. The abnormality detection method as described in any one of these. 26. Determine the representative position on the input image of the group of people whose detection of the state change was the basis for the abnormal state determination, Causing the output device to output an image in which the input image is colored in a color corresponding to the distance from the representative position in a circle or ellipse centered on the representative position; Further includes: To 25. The abnormality detection method as described in any one of these. 27.
- the output of the image is continued for a duration corresponding to the number of persons in the group of persons whose detection of the state change is the basis for the abnormal state determination.
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Signal Processing (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Human Computer Interaction (AREA)
- Psychiatry (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Social Psychology (AREA)
- Emergency Management (AREA)
- Business, Economics & Management (AREA)
- Image Analysis (AREA)
- Alarm Systems (AREA)
- Closed-Circuit Television Systems (AREA)
Abstract
Description
〔装置構成〕
図1は、第一実施形態における画像処理装置(以降、単に処理装置と表記する場合もある)1のハードウェア構成例を概念的に示す図である。処理装置1は、いわゆるコンピュータであり、例えば、バスで相互に接続される、CPU(Central Processing Unit)2、メモリ3、通信ユニット4、入出力インタフェース(I/F)5等を有する。メモリ3は、RAM(Random Access Memory)、ROM(Read Only Memory)、ハードディスク等である。通信ユニット4は、他のコンピュータや機器と信号のやりとりを行う。第一実施形態では、通信ユニット4には監視カメラ7が接続される。通信ユニット4は、監視カメラ7から映像信号を取得する。また、通信ユニット4には、可搬型記録媒体等も接続され得る。
図2は、第一実施形態における画像処理装置1の処理構成例を概念的に示す図である。図2に示されるように、処理装置1は、取得部11、人物検出部12、変化検出部13、保持部14、判定部15、出力処理部16等を有する。これら各処理部は、例えば、CPU2によりメモリ3に格納されるプログラムが実行されることにより実現される。また、当該プログラムは、例えば、CD(Compact Disc)、メモリカード等のような可搬型記録媒体やネットワーク上の他のコンピュータから通信ユニット4又は入出力I/F5を介してインストールされ、メモリ3に格納されてもよい。
判定部15は、複数人での対象の状態変化の発生時期のずれを異常状態の判定のために更に用いることもできる。この場合、判定部15は、対象の状態変化が所定人数以上で生じていることを検知すると、その状態変化の発生時期の集合の時間幅を計算する。判定部15は、その時間幅が所定時間幅内か否かを判定する。即ち、判定部15は、所定時間内で対象の状態変化が複数人で生じたことを検出し、この検出結果に応じて異常状態を判定する。複数人での対象の状態変化の発生時期のずれが小さい程、何らかの異常状態が発生している可能性が高い。ほぼ同じタイミングで複数人が対象の状態変化を生じさせることは、異常状態の発生に起因する可能性が高いからである。このように、「所定時間内」という条件を更に設けて異常状態を判定することで、異常状態の判定精度を向上させることができる。
判定部15は、変化後の状態が同一の上記対象の状態変化が複数人で生じたことを検出し、この検出に応じて異常状態を判定することもできる。この場合、判定部15は、保持部14に保持される状態変化情報の中で、状態変化の内容が同一の人物の数をカウントし、この人数が所定人数以上となる状態変化が存在した場合に、異常状態有りと判定する。複数人が同じ状態変化を起こした場合には、何かしらの異常状態が生じている可能性が高い。よって、このように「変化後の状態が同一」という条件を異常状態判定に更に加えることで、異常状態の判定精度を高めることができる。
判定部15は、対象の状態変化が距離的に近い複数人で生じたことを検出し、この検出に応じて異常状態を判定することもできる。この場合、判定部15は、保持部14に保持される状態変化情報の位置情報を用いて、対象の状態変化を起こした各人の相対的距離(画像内距離)を算出する。判定部15は、算出された相対距離が所定距離以内となる人数をカウントし、その人数が所定人数以上となる場合に、異常状態有りと判定する。距離的に近い複数人が対象の状態変化を起こした場合には、何かしらの異常状態が生じている可能性が高い。よって、このように「対象の状態変化を起こした各人の相対距離が近い」という条件を異常状態判定に更に加えることで、異常状態の判定精度を高めることができる。
出力処理部16は、入力画像中の、状態変化の検出が異常状態判定の根拠とされた人物群又は当該人物群の各人に、描画要素が付された画像を出力装置6に出力させる。描画要素を付す位置は、例えば、保持部14に保持される状態変化情報の位置情報により決定される。以降、状態変化の検出が異常状態判定の根拠とされた人物群を「根拠となる人物群」と表記する場合もある。出力処理部16は、保持部14に保持される入力画像に対して当該描画要素を付して得られる画像を出力装置6に出力させる。当該描画要素は、例えば、根拠となる人物群を囲む線、その人物群又はその人物群の各人に付される何らかのマーク(点、星、感嘆符、矢印、吹き出し等)である。根拠となる人物群又はその各人が特定できるように付されていれば、当該描画要素自体及び当該描画要素を付す位置及びその付し方は、制限されない。
出力処理部16は、入力画像中の、状態変化の検出が異常状態判定の根拠とされた人物群に対して、当該人物群の人数及び当該人物群の各人の相対距離のいずれか一方又は両方が更に付された画像を出力装置6に出力させることもできる。人数及び相対距離は、対象の人物群を特定可能な位置に付される。ここで、根拠となる人物群の人数及び当該人物群の各人の相対距離は、異常状態の判定精度を表すと考えられる。対象の状態変化を起こした人数が多い場合、及び、対象の状態変化を起こした各人の距離が近い場合には、異常状態が存在している可能性が高いからである。よって、このような出力画像により、その画像を見た者が、異常状態の判定精度を容易に推測できるようになる。
出力処理部16は、上記描画要素として又は上記描画要素と共に、状態変化の検出が異常状態判定の根拠とされた人物群の各人に、状態変化の順番又は状態変化の時間差が付された画像を出力装置6に出力させることもできる。各人物の状態変化の時間としては、その状態変化が検出された各入力画像に対応する時間が用いられる。ここで、状態変化の時間差とは、最先の状態変化が生じた時間から、各人が起こした状態変化の時間までの間の時間差である。例えば、最先の状態変化を起こした人の周囲に「±0」が付され、その他の各人の周囲にその人が起こした状態変化の時間までの時間差「+5秒」、「+8秒」が付される。異常状態に伴う状態変化は、その異常状態に近い者から起こす可能性が高い。よって、状態変化の順番又は状態変化の時間差を付した画像を出力することにより、その画像を見た者が、異常状態の存在場所を容易に推測できるようになる。
出力処理部16は、異常状態判定の根拠とされた複数人の状態変化における変化後の状態に対応する方向を表す描画要素が付された画像を出力装置6に出力させることもできる。この方向を表す描画要素は、根拠とされた人物群又は当該人物群の各人に付される。出力処理部16は、変化後の状態と対応する方向情報との関連付けを予め保持する。例えば、変化後の状態が「しゃがむ」の場合、異常状態は、その人の上方又は下方に存在する可能性が高いため、変化後の状態「しゃがむ」に対して「上方向」及び「下方向」が対応付けられる。また、状態変化が「駆け出す」の場合、異常状態は、その駆け出した方向とは逆方向に存在する可能性が高いため、「駆け出す」に対応する変化後の状態に対して「駆け出す方向とは逆方向」が対応付けられる。出力処理部16は、このような関連付け情報を用いて、変化後の状態に対応する方向情報を特定し、この特定された方向情報に対応する描画要素を選択する。方向を表す描画要素には、例えば、矢印等が挙げられる。
出力処理部16は、根拠となる人物群の入力画像上での代表位置を中心とする円状又は楕円状に、入力画像がその代表位置からの距離に対応する色で着色された画像を出力装置6に出力させることもできる。この場合、処理装置1は、次のような処理構成を有すればよい。
以下、第一実施形態における異常検知方法について図11を用いて説明する。図11は、第一実施形態における処理装置1の動作例を示すフローチャートである。図11に示されるように、第一実施形態における異常検知方法は、処理装置1のような少なくとも1つのコンピュータにより実行される。例えば、図示される各工程は、処理装置1が有する各処理部により実行される。各工程は、処理装置1が有する上述の各処理部の処理内容と同様であるため、各工程の詳細は、適宜省略される。
上述のように、第一実施形態では、入力画像に基づいて、人物における対象の状態変化が検出され、対象の状態変化が所定人数以上で生じた場合に、異常状態有りと判定される。人物の不審な状態のような人物の異常状態、放置物、汚物等のような物体の異常状態、火災や事故のような場の異常状態が発生した場合、その発生場所の周囲に人が存在している場合には、その人々は、その監視対象場所において通常時には起こさない姿勢、行動、表情等の見た目の変化を起こす。そこで、第一実施形態では、人物、物体、場等における様々な異常状態がその周囲の人々の状態変化により捉えられる。
第二実施形態における処理装置1は、異常状態の判定結果を用いて、監視カメラ7を制御する。以下、第二実施形態における処理装置1について、第一実施形態と異なる内容を中心に説明する。以下の説明では、第一実施形態と同様の内容については適宜省略する。
制御部18は、状態変化の検出が異常状態判定の根拠とされた人物群の入力画像上での代表位置に基づいて、監視カメラ7の撮像方向を制御することができる。この場合、図10に示されるように、処理装置1は、決定部17を更に有する。決定部17は、上述のとおり、状態変化の検出が異常状態判定の根拠とされた人物群の入力画像上での代表位置を決定する。
制御部18は、状態変化の検出が異常状態判定の根拠とされた人物群の入力画像上での大きさに基づいて、監視カメラ7の画角を制御することができる。例えば、制御部18は、根拠となる人物群が入力画像内の一部の所定範囲に収まり、その人物群の周辺が入力画像に映るように、監視カメラ7の画角を変更する指示を監視カメラ7に送る。この指示に応じて、監視カメラ7は、ズームイン又はズームアウトする。人物群の入力画像上の大きさは、例えば、人物検出部12により検出された人物の画像領域を足し合わせることで決めることができる。制御部18は、当該人物群の大きさが当該所定範囲内に収まるまで、ズームイン又はズームアウトする指示を監視カメラ7に送り続けてもよい。
制御部18は、状態変化の検出が異常状態判定の根拠とされた人物群の全体又は一部が入力画像に含まれる範囲で、監視カメラ7の撮像方向を連続的に変化させることができる。具体的には、制御部18は、保持部14に保持される状態変化情報に基づいて、根拠となる人物群の入力画像上の位置を認識する。制御部18は、この認識された位置に基づいて、当該人物群の少なくとも一部が入力画像に収まる範囲に対応する、撮像方向の変更範囲を算出することができる。制御部18は、この変更範囲内で、当該人物群の入力画像上の位置が変わるように、撮像方向を変更する指示を連続的に監視カメラ7又は撮像方向を調整する機構に送る。制御部18は、例えば、撮像方向の変更指示に、変更すべき方向(上下左右等)及び変更すべき角度を含める。
制御部18は、異常状態判定の根拠とされた複数人の状態変化における変化後の状態に対応する方向に監視カメラ7の撮像方向を変更させることもできる。変化後の状態に対応する方向、及び、その方向情報の取得手法については、《出力例4》で説明したとおりである。制御部18は、変化後の状態に対応する方向情報を取得し、その方向情報を含む撮像方向変更指示を監視カメラ7又は撮像方向を調整する機構に送る。
以下、第二実施形態における異常検知方法について図13を用いて説明する。図13は、第二実施形態における処理装置1の動作例を示すフローチャートである。図13では、図11と同じ処理内容の工程には、図11と同じ符号が付されている。第二実施形態における異常検知方法の実行主体は、第一実施形態と同様である。
上述のとおり、第二実施形態では、異常状態有りと判定された場合に、根拠となる人物群の位置やその人物群の各人の変化後の状態等に基づいて、監視カメラ7の撮像方向や画角が制御される。結果として、異常状態が存在している可能性の高い、根拠となる人物群の周囲の状況が視認可能となるため、異常状態を確認し易くすることができる。
処理装置1は、複数の監視カメラ7に接続されており、当該複数の監視カメラ7を制御することもできる。この変形例では、各監視カメラ7は、異なる監視対象場所が撮影範囲となるように、それぞれ設置される。この変形例では、処理装置1は、或る監視カメラ7で撮像された入力画像に基づいて異常状態が検知された場合に、その異常状態判定の根拠とされた人物群の位置に対応する他の監視カメラ7を選択し、その監視カメラ7の撮像方向及び画角を変更させてもよい。このようにすれば、1台の監視カメラ7の撮像方向や画角を上述のように変更しても、その監視カメラ7では映すことができない位置に異常状態が存在する場合でも、周囲の他の監視カメラ7を制御することで、その位置を撮像することができる。
処理装置1には監視カメラ7が接続されず、処理装置1が他のコンピュータに接続される監視カメラ7を用いてもよい。この場合、取得部11は、可搬型記録媒体、又は、他のコンピュータから通信ユニット4を介して、入力画像又は映像信号を取得すればよい。当該他のコンピュータとしては、例えば、カメラで撮影した画像を一旦蓄積して配信する、画像蓄積配信装置や、画像を蓄積し、再生する画像録画器等があり得る。この場合、出力処理部16は、その画像を他のコンピュータに対して通信ユニット4を介して出力してもよい。また、制御部18は、当該他のコンピュータに監視カメラ7を制御するための指示情報を送信する。
以下、第三実施形態における画像処理装置及び異常検知方法について図14及び図15を用いて説明する。
前記対象の状態変化が複数人で生じたことを検出し、該検出に応じて異常状態を判定する判定手段と、
を備える画像処理装置。
2. 前記判定手段は、変化後の状態が同一の前記対象の状態変化が複数人で生じたことを検出する、
1.に記載の画像処理装置。
3. 前記判定手段は、前記対象の状態変化が距離的に近い複数人で生じたことを検出する、
1.又は2.に記載の画像処理装置。
4. 前記判定手段は、前記複数人での前記対象の状態変化の発生時期のずれを異常状態の判定のために更に用いる、
1.から3.のいずれか1つに記載の画像処理装置。
5. 前記判定手段は、前記複数人での前記対象の状態変化の発生時期のずれに応じて、異常状態の程度を推測する、
4.に記載の画像処理装置。
6. 入力画像中の、状態変化の検出が異常状態判定の根拠とされた人物群又は該人物群の各人に、描画要素が付された画像を出力装置に出力させる出力処理手段
を更に備える1.から5.のいずれか1つに記載の画像処理装置。
7. 前記出力処理手段は、入力画像中の、状態変化の検出が異常状態判定の根拠とされた人物群に対して、該人物群の人数及び該人物群の各人の相対距離のいずれか一方又は両方が更に付された前記画像を前記出力装置に出力させる、
6.に記載の画像処理装置。
8. 前記出力処理手段は、前記描画要素として又は前記描画要素と共に、状態変化の検出が異常状態判定の根拠とされた人物群の各人に、状態変化の順番又は状態変化の時間差が付された前記画像を前記出力装置に出力させる、
6.又は7.に記載の画像処理装置。
9. 前記出力処理手段は、異常状態判定の根拠とされた複数人の状態変化における変化後の状態に対応する方向を表す前記描画要素が付された前記画像を前記出力装置に出力させる、
6.から8.のいずれか1つに記載の画像処理装置。
10. 状態変化の検出が異常状態判定の根拠とされた人物群の入力画像上での代表位置を決定する決定手段と、
前記代表位置を中心とする円状又は楕円状に、入力画像が前記代表位置からの距離に対応する色で着色された画像を出力装置に出力させる出力処理手段と、
を更に備える1.から9.のいずれか1つに記載の画像処理装置。
11. 前記出力処理手段は、状態変化の検出が異常状態判定の根拠とされた人物群の人数に対応する継続時間で、前記画像の出力を継続させる、
6.から10.のいずれか1つに記載の画像処理装置。
12. 状態変化の検出が異常状態判定の根拠とされた人物群の入力画像上での代表位置を決定する決定手段と、
前記代表位置に基づいて、撮像装置の撮像方向を制御する制御手段と、
を更に備える1.から11.のいずれか1つに記載の画像処理装置。
13. 状態変化の検出が異常状態判定の根拠とされた人物群の入力画像上での大きさに基づいて、撮像装置の画角を制御する制御手段、
を更に備える1.から12.のいずれか1つに記載の画像処理装置。
14. 状態変化の検出が異常状態判定の根拠とされた人物群の全体又は一部が入力画像に含まれる範囲で、撮像装置の撮像方向を連続的に変化させる制御手段、
を更に備える1.から13.のいずれか1つに記載の画像処理装置。
15. 異常状態判定の根拠とされた複数人の状態変化における変化後の状態に対応する方向に撮像装置の撮像方向を変更させる制御手段、
を更に備える1.から14.のいずれか1つに記載の画像処理装置。
16. 前記制御手段は、状態変化の検出が異常状態判定の根拠とされた人物群の人数に対応する継続時間で、前記撮像装置の制御を継続する、
12.から15.のいずれか1つに記載の画像処理装置。
入力画像に基づいて人物における対象の状態変化を検出し、
前記対象の状態変化が複数人で生じたことを検出し、
前記検出に応じて異常状態を判定する、
ことを含む異常検知方法。
18. 前記対象の状態変化が複数人で生じたことの検出は、変化後の状態が同一の前記対象の状態変化が複数人で生じたことを検出する、
17.に記載の異常検知方法。
19. 前記対象の状態変化が複数人で生じたことの検出は、前記対象の状態変化が距離的に近い複数人で生じたことを検出する、
17.又は18.に記載の異常検知方法。
20. 前記対象の状態変化が複数人で生じたことの検出は、前記複数人での前記対象の状態変化の発生時期のずれが所定時間内であることを更に判断する、
17.から19.のいずれか1つに記載の異常検知方法。
21. 前記複数人での前記対象の状態変化の発生時期のずれに応じて、異常状態の程度を推測する、
ことを更に含む20.に記載の異常検知方法。
22. 入力画像中の、状態変化の検出が異常状態判定の根拠とされた人物群又は該人物群の各人に、描画要素が付された画像を出力装置に出力させる
ことを更に含む17.から21.のいずれか1つに記載の異常検知方法。
23. 前記出力は、入力画像中の、状態変化の検出が異常状態判定の根拠とされた人物群に対して、該人物群の人数及び該人物群の各人の相対距離のいずれか一方又は両方が更に付された前記画像を前記出力装置に出力させる、
22.に記載の異常検知方法。
24. 前記出力は、前記描画要素として又は前記描画要素と共に、状態変化の検出が異常状態判定の根拠とされた人物群の各人に、状態変化の順番又は状態変化の時間差が付された前記画像を前記出力装置に出力させる、
22.又は23.に記載の異常検知方法。
25. 前記出力は、異常状態判定の根拠とされた複数人の状態変化における変化後の状態に対応する方向を表す前記描画要素が付された前記画像を前記出力装置に出力させる、
22.から24.のいずれか1つに記載の異常検知方法。
26. 状態変化の検出が異常状態判定の根拠とされた人物群の入力画像上での代表位置を決定し、
前記代表位置を中心とする円状又は楕円状に、入力画像が前記代表位置からの距離に対応する色で着色された画像を出力装置に出力させる、
ことを更に含む17.から25.のいずれか1つに記載の異常検知方法。
27. 状態変化の検出が異常状態判定の根拠とされた人物群の人数に対応する継続時間で、前記画像の出力を継続させる、
ことを更に含む22.から26.のいずれか1つに記載の異常検知方法。
28. 状態変化の検出が異常状態判定の根拠とされた人物群の入力画像上での代表位置を決定し、
前記代表位置に基づいて、撮像装置の撮像方向を制御する、
ことを更に含む17.から27.のいずれか1つに記載の異常検知方法。
29. 状態変化の検出が異常状態判定の根拠とされた人物群の入力画像上での大きさに基づいて、撮像装置の画角を制御する、
ことを更に含む17.から28.のいずれか1つに記載の異常検知方法。
30. 状態変化の検出が異常状態判定の根拠とされた人物群の全体又は一部が入力画像に含まれる範囲で、撮像装置の撮像方向を連続的に変化させる、
ことを更に含む17.から29.のいずれか1つに記載の異常検知方法。
31. 異常状態判定の根拠とされた複数人の状態変化における変化後の状態に対応する方向に撮像装置の撮像方向を変更させる、
ことを更に含む17.から30.のいずれか1つに記載の異常検知方法。
32. 状態変化の検出が異常状態判定の根拠とされた人物群の人数に対応する継続時間で、前記撮像装置の制御を継続する、
ことを更に含む28.から31.のいずれか1つに記載の異常検知方法。
34. 33.に記載のプログラムをコンピュータに読み取り可能に記録する記録媒体。
Claims (19)
- 入力画像に基づいて人物における対象の状態変化を検出する変化検出手段と、
前記対象の状態変化が複数人で生じたことを検出し、該検出に応じて異常状態を判定する判定手段と、
を備える画像処理装置。 - 前記対象の状態変化は、異常状態の発生に起因して生じる人物の見た目の変化である、
請求項1に記載の画像処理装置。 - 前記判定手段は、変化後の状態が同一の前記対象の状態変化が複数人で生じたことを検出する、
請求項1又は2に記載の画像処理装置。 - 前記判定手段は、前記対象の状態変化が距離的に近い複数人で生じたことを検出する、
請求項1から3のいずれか1項に記載の画像処理装置。 - 前記判定手段は、前記複数人での前記対象の状態変化の発生時期のずれを異常状態の判定のために更に用いる、
請求項1から4のいずれか1項に記載の画像処理装置。 - 前記判定手段は、前記複数人での前記対象の状態変化の発生時期のずれに応じて、異常状態の程度を推測する、
請求項5に記載の画像処理装置。 - 入力画像中の、状態変化の検出が異常状態判定の根拠とされた人物群又は該人物群の各人に、描画要素が付された画像を出力装置に出力させる出力処理手段
を更に備える請求項1から6のいずれか1項に記載の画像処理装置。 - 前記出力処理手段は、入力画像中の、状態変化の検出が異常状態判定の根拠とされた人物群に対して、該人物群の人数及び該人物群の各人の相対距離のいずれか一方又は両方が更に付された前記画像を前記出力装置に出力させる、
請求項7に記載の画像処理装置。 - 前記出力処理手段は、前記描画要素として又は前記描画要素と共に、状態変化の検出が異常状態判定の根拠とされた人物群の各人に、状態変化の順番又は状態変化の時間差が付された前記画像を前記出力装置に出力させる、
請求項7又は8に記載の画像処理装置。 - 前記出力処理手段は、異常状態判定の根拠とされた複数人の状態変化における変化後の状態に対応する方向を表す前記描画要素が付された前記画像を前記出力装置に出力させる、
請求項7から9のいずれか1項に記載の画像処理装置。 - 状態変化の検出が異常状態判定の根拠とされた人物群の入力画像上での代表位置を決定する決定手段と、
前記代表位置を中心とする円状又は楕円状に、入力画像が前記代表位置からの距離に対応する色で着色された画像を出力装置に出力させる出力処理手段と、
を更に備える請求項1から10のいずれか1項に記載の画像処理装置。 - 前記出力処理手段は、状態変化の検出が異常状態判定の根拠とされた人物群の人数に対応する継続時間で、前記画像の出力を継続させる、
請求項7から11のいずれか1項に記載の画像処理装置。 - 状態変化の検出が異常状態判定の根拠とされた人物群の入力画像上での代表位置を決定する決定手段と、
前記代表位置に基づいて、撮像装置の撮像方向を制御する制御手段と、
を更に備える請求項1から12のいずれか1項に記載の画像処理装置。 - 状態変化の検出が異常状態判定の根拠とされた人物群の入力画像上での大きさに基づいて、撮像装置の画角を制御する制御手段、
を更に備える請求項1から13のいずれか1項に記載の画像処理装置。 - 状態変化の検出が異常状態判定の根拠とされた人物群の全体又は一部が入力画像に含まれる範囲で、撮像装置の撮像方向を連続的に変化させる制御手段、
を更に備える請求項1から14のいずれか1項に記載の画像処理装置。 - 異常状態判定の根拠とされた複数人の状態変化における変化後の状態に対応する方向に撮像装置の撮像方向を変更させる制御手段、
を更に備える請求項1から15のいずれか1項に記載の画像処理装置。 - 前記制御手段は、状態変化の検出が異常状態判定の根拠とされた人物群の人数に対応する継続時間で、前記撮像装置の制御を継続する、
請求項13から16のいずれか1項に記載の画像処理装置。 - 少なくとも1つのコンピュータにより実行される異常検知方法において、
入力画像に基づいて人物における対象の状態変化を検出し、
前記対象の状態変化が複数人で生じたことを検出し、
前記検出に応じて異常状態を判定する、
ことを含む異常検知方法。 - 請求項18に記載の異常検知方法をコンピュータに実行させるプログラム。
Priority Applications (6)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US15/321,640 US10846536B2 (en) | 2014-06-27 | 2015-05-25 | Abnormality detection device and abnormality detection method |
JP2016529191A JP6696422B2 (ja) | 2014-06-27 | 2015-05-25 | 異常検知装置及び異常検知方法 |
US16/298,746 US11106918B2 (en) | 2014-06-27 | 2019-03-11 | Abnormality detection device and abnormality detection method |
US16/298,698 US11250268B2 (en) | 2014-06-27 | 2019-03-11 | Abnormality detection device and abnormality detection method |
US16/862,382 US20200257905A1 (en) | 2014-06-27 | 2020-04-29 | Abnormality detection device and abnormality detection method |
US18/229,945 US20230410513A1 (en) | 2014-06-27 | 2023-08-03 | Abnormality detection device and abnormality detection method |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2014132304 | 2014-06-27 | ||
JP2014-132304 | 2014-06-27 |
Related Child Applications (3)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US15/321,640 A-371-Of-International US10846536B2 (en) | 2014-06-27 | 2015-05-25 | Abnormality detection device and abnormality detection method |
US16/298,746 Continuation US11106918B2 (en) | 2014-06-27 | 2019-03-11 | Abnormality detection device and abnormality detection method |
US16/298,698 Continuation US11250268B2 (en) | 2014-06-27 | 2019-03-11 | Abnormality detection device and abnormality detection method |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2015198767A1 true WO2015198767A1 (ja) | 2015-12-30 |
Family
ID=54937864
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2015/064866 WO2015198767A1 (ja) | 2014-06-27 | 2015-05-25 | 異常検知装置及び異常検知方法 |
Country Status (3)
Country | Link |
---|---|
US (5) | US10846536B2 (ja) |
JP (4) | JP6696422B2 (ja) |
WO (1) | WO2015198767A1 (ja) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20170220871A1 (en) * | 2014-06-27 | 2017-08-03 | Nec Corporation | Abnormality detection device and abnormality detection method |
JP6247786B1 (ja) * | 2017-04-06 | 2017-12-13 | イセット株式会社 | 自動通報装置 |
JP2018082281A (ja) * | 2016-11-15 | 2018-05-24 | キヤノン株式会社 | 情報処理装置、情報処理装置の制御方法およびプログラム |
JPWO2018025831A1 (ja) * | 2016-08-04 | 2019-06-13 | 日本電気株式会社 | 人流推定装置、人流推定方法およびプログラム |
WO2020049981A1 (en) * | 2018-09-05 | 2020-03-12 | Nec Corporation | An apparatus and a method for adaptively managing event-related data in a control room |
Families Citing this family (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108647631B (zh) * | 2013-06-28 | 2023-04-07 | 日本电气株式会社 | 人群状态识别设备、方法和计算机可读记录介质 |
US10244175B2 (en) * | 2015-03-09 | 2019-03-26 | Apple Inc. | Automatic cropping of video content |
JP6835065B2 (ja) | 2016-03-18 | 2021-02-24 | 日本電気株式会社 | 情報処理装置、制御方法、及びプログラム |
EP3566175A4 (en) * | 2017-01-06 | 2020-07-29 | Sportlogiq Inc. | SYSTEMS AND METHODS FOR UNDERSTANDING BEHAVIOR FROM TRAJECTORIES |
JP6469139B2 (ja) * | 2017-01-17 | 2019-02-13 | キヤノン株式会社 | 情報処理装置、情報処理方法及びプログラム |
US11175650B2 (en) * | 2017-11-03 | 2021-11-16 | Drishti Technologies, Inc. | Product knitting systems and methods |
JP7104523B2 (ja) * | 2018-02-05 | 2022-07-21 | キヤノン株式会社 | 情報処理装置、システム、情報処理装置の制御方法、及び、プログラム |
CN109359510B (zh) * | 2018-08-27 | 2022-03-22 | 广州大学 | 一种对异常行为的视觉分析处理方法 |
US10824868B2 (en) * | 2018-09-27 | 2020-11-03 | Ncr Corporation | Image processing for determining relationships between tracked objects |
US11055539B2 (en) * | 2018-09-27 | 2021-07-06 | Ncr Corporation | Image processing for distinguishing individuals in groups |
JP2020072457A (ja) * | 2018-11-02 | 2020-05-07 | ソニー株式会社 | 映像処理装置、映像処理方法、およびプログラム |
JP7201072B2 (ja) * | 2019-03-27 | 2023-01-10 | 日本電気株式会社 | 監視装置、不審オブジェクト検出方法、およびプログラム |
US11151858B2 (en) * | 2020-06-24 | 2021-10-19 | Paul Edward Moshe Modiano | Crowd management in an enclosed premises |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2002021441A1 (fr) * | 2000-09-06 | 2002-03-14 | Hitachi, Ltd. | Detecteur de comportement anormal |
JP2007243342A (ja) * | 2006-03-06 | 2007-09-20 | Yokogawa Electric Corp | 画像監視装置及び画像監視システム |
JP2010128594A (ja) * | 2008-11-25 | 2010-06-10 | Sogo Keibi Hosho Co Ltd | 動作検出装置および動作検出方法 |
JP2012208850A (ja) * | 2011-03-30 | 2012-10-25 | Secom Co Ltd | 異常検知装置 |
Family Cites Families (46)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH06266840A (ja) | 1993-03-11 | 1994-09-22 | Hitachi Ltd | 移動物体の状態検出装置 |
US5831669A (en) * | 1996-07-09 | 1998-11-03 | Ericsson Inc | Facility monitoring system with image memory and correlation |
US5747719A (en) * | 1997-01-21 | 1998-05-05 | Bottesch; H. Werner | Armed terrorist immobilization (ATI) system |
JPH10285586A (ja) | 1997-04-02 | 1998-10-23 | Omron Corp | 画像処理装置、監視画像表示システム、画像処理方法、および、監視画像表示方法 |
US6147601A (en) | 1999-01-09 | 2000-11-14 | Heat - Timer Corp. | Electronic message delivery system utilizable in the monitoring of remote equipment and method of same |
US6678413B1 (en) | 2000-11-24 | 2004-01-13 | Yiqing Liang | System and method for object identification and behavior characterization using video analysis |
JP2002344946A (ja) | 2001-05-16 | 2002-11-29 | Oki Electric Ind Co Ltd | 監視装置 |
US6975346B2 (en) | 2002-06-27 | 2005-12-13 | International Business Machines Corporation | Method for suspect identification using scanning of surveillance media |
US7999857B2 (en) * | 2003-07-25 | 2011-08-16 | Stresscam Operations and Systems Ltd. | Voice, lip-reading, face and emotion stress analysis, fuzzy logic intelligent camera system |
JP4368767B2 (ja) * | 2004-09-08 | 2009-11-18 | 独立行政法人産業技術総合研究所 | 異常動作検出装置および異常動作検出方法 |
US7606425B2 (en) | 2004-09-09 | 2009-10-20 | Honeywell International Inc. | Unsupervised learning of events in a video sequence |
WO2006080367A1 (ja) | 2005-01-28 | 2006-08-03 | Olympus Corporation | 粒子群運動解析システム、粒子群運動解析方法及びプログラム |
US20060190419A1 (en) * | 2005-02-22 | 2006-08-24 | Bunn Frank E | Video surveillance data analysis algorithms, with local and network-shared communications for facial, physical condition, and intoxication recognition, fuzzy logic intelligent camera system |
JP4622702B2 (ja) | 2005-05-27 | 2011-02-02 | 株式会社日立製作所 | 映像監視装置 |
EP1748378B1 (en) | 2005-07-26 | 2009-09-16 | Canon Kabushiki Kaisha | Image capturing apparatus and image capturing method |
JP2007264950A (ja) | 2006-03-28 | 2007-10-11 | Toyota Motor Corp | 自律移動型ロボット |
JP2007272488A (ja) * | 2006-03-31 | 2007-10-18 | Yokogawa Electric Corp | 画像処理装置、監視カメラ及び画像監視システム |
US8269834B2 (en) | 2007-01-12 | 2012-09-18 | International Business Machines Corporation | Warning a user about adverse behaviors of others within an environment based on a 3D captured image stream |
US20080252722A1 (en) * | 2007-04-11 | 2008-10-16 | Yuan-Kai Wang | System And Method Of Intelligent Surveillance And Analysis |
US7965866B2 (en) * | 2007-07-03 | 2011-06-21 | Shoppertrak Rct Corporation | System and process for detecting, tracking and counting human objects of interest |
US8472715B2 (en) * | 2007-10-26 | 2013-06-25 | Panasonic Corporation | Situation determining apparatus, situation determining method, situation determining program, abnormality determining apparatus, abnormality determining method, abnormality determining program, and congestion estimating apparatus |
WO2011114770A1 (ja) * | 2010-03-15 | 2011-09-22 | オムロン株式会社 | 監視カメラ端末 |
JP2011198161A (ja) | 2010-03-22 | 2011-10-06 | Hiromitsu Hama | 対象物認識システム及び該システムを利用する監視システム、見守りシステム |
JP5473801B2 (ja) | 2010-06-30 | 2014-04-16 | 株式会社日立製作所 | 監視装置 |
JP5400718B2 (ja) | 2010-07-12 | 2014-01-29 | 株式会社日立国際電気 | 監視システムおよび監視方法 |
JP5435740B2 (ja) | 2010-11-12 | 2014-03-05 | オムロン株式会社 | 監視システム、および監視サーバ |
JP5810296B2 (ja) * | 2011-06-07 | 2015-11-11 | パナソニックIpマネジメント株式会社 | 画像表示装置及び画像表示方法 |
KR101179276B1 (ko) | 2011-06-13 | 2012-09-03 | 고려대학교 산학협력단 | 비정상 집단 행동 탐지 방법 및 장치 |
JP5560397B2 (ja) | 2011-12-22 | 2014-07-23 | 株式会社ウェルソック | 自律型防犯警戒システム及び自律型防犯警戒方法 |
US8798318B2 (en) | 2012-01-18 | 2014-08-05 | Xerox Corporation | System and method for video episode viewing and mining |
TWI582707B (zh) * | 2012-01-19 | 2017-05-11 | Utechzone Co Ltd | 車內眼控的方法 |
KR101346773B1 (ko) | 2012-03-09 | 2013-12-31 | 고려대학교 산학협력단 | H.264 압축영역에서의 비정상 군중행동 탐지 및 분석 방법 |
US10368010B2 (en) * | 2012-07-31 | 2019-07-30 | Nec Corporation | Image processing system, image processing method, and program |
US20140267736A1 (en) | 2013-03-15 | 2014-09-18 | Bruno Delean | Vision based system for detecting a breach of security in a monitored location |
US20140362213A1 (en) | 2013-06-05 | 2014-12-11 | Vincent Tseng | Residence fall and inactivity monitoring system |
JP5632512B1 (ja) | 2013-07-02 | 2014-11-26 | パナソニック株式会社 | 人物行動分析装置、人物行動分析システムおよび人物行動分析方法、ならびに監視装置 |
US9159218B2 (en) * | 2013-09-17 | 2015-10-13 | Microchip Technology Incorporated | Initiation of carbon monoxide and/or smoke detector alarm test using image recognition and/or facial gesturing |
JP6297822B2 (ja) * | 2013-11-19 | 2018-03-20 | ルネサスエレクトロニクス株式会社 | 検知装置、検知システム及び検知方法 |
US10846536B2 (en) * | 2014-06-27 | 2020-11-24 | Nec Corporation | Abnormality detection device and abnormality detection method |
US9449218B2 (en) | 2014-10-16 | 2016-09-20 | Software Ag Usa, Inc. | Large venue surveillance and reaction systems and methods using dynamically analyzed emotional input |
WO2017020142A1 (zh) | 2015-08-03 | 2017-02-09 | 深圳市好航科技有限公司 | 多用途车辆智能监控***及方法 |
US9615066B1 (en) | 2016-05-03 | 2017-04-04 | Bao Tran | Smart lighting and city sensor |
KR20180060784A (ko) | 2016-11-29 | 2018-06-07 | 삼성전자주식회사 | 비정상 객체 판단 방법 및 장치 |
US10186124B1 (en) * | 2017-10-26 | 2019-01-22 | Scott Charles Mullins | Behavioral intrusion detection system |
US10043360B1 (en) * | 2017-10-26 | 2018-08-07 | Scott Charles Mullins | Behavioral theft detection and notification system |
WO2020088739A1 (en) * | 2018-10-29 | 2020-05-07 | Hexagon Technology Center Gmbh | Facility surveillance systems and methods |
-
2015
- 2015-05-25 US US15/321,640 patent/US10846536B2/en active Active
- 2015-05-25 WO PCT/JP2015/064866 patent/WO2015198767A1/ja active Application Filing
- 2015-05-25 JP JP2016529191A patent/JP6696422B2/ja active Active
-
2019
- 2019-03-11 US US16/298,746 patent/US11106918B2/en active Active
- 2019-03-11 US US16/298,698 patent/US11250268B2/en active Active
-
2020
- 2020-01-16 JP JP2020005225A patent/JP6863488B2/ja active Active
- 2020-04-29 US US16/862,382 patent/US20200257905A1/en active Pending
-
2021
- 2021-03-25 JP JP2021051030A patent/JP7173196B2/ja active Active
-
2022
- 2022-11-02 JP JP2022175989A patent/JP7396436B2/ja active Active
-
2023
- 2023-08-03 US US18/229,945 patent/US20230410513A1/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2002021441A1 (fr) * | 2000-09-06 | 2002-03-14 | Hitachi, Ltd. | Detecteur de comportement anormal |
JP2007243342A (ja) * | 2006-03-06 | 2007-09-20 | Yokogawa Electric Corp | 画像監視装置及び画像監視システム |
JP2010128594A (ja) * | 2008-11-25 | 2010-06-10 | Sogo Keibi Hosho Co Ltd | 動作検出装置および動作検出方法 |
JP2012208850A (ja) * | 2011-03-30 | 2012-10-25 | Secom Co Ltd | 異常検知装置 |
Cited By (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20170220871A1 (en) * | 2014-06-27 | 2017-08-03 | Nec Corporation | Abnormality detection device and abnormality detection method |
US20190205657A1 (en) * | 2014-06-27 | 2019-07-04 | Nec Corporation | Abnormality detection device and abnormality detection method |
US10846536B2 (en) * | 2014-06-27 | 2020-11-24 | Nec Corporation | Abnormality detection device and abnormality detection method |
US11106918B2 (en) | 2014-06-27 | 2021-08-31 | Nec Corporation | Abnormality detection device and abnormality detection method |
US11250268B2 (en) | 2014-06-27 | 2022-02-15 | Nec Corporation | Abnormality detection device and abnormality detection method |
JPWO2018025831A1 (ja) * | 2016-08-04 | 2019-06-13 | 日本電気株式会社 | 人流推定装置、人流推定方法およびプログラム |
JP2018082281A (ja) * | 2016-11-15 | 2018-05-24 | キヤノン株式会社 | 情報処理装置、情報処理装置の制御方法およびプログラム |
JP6247786B1 (ja) * | 2017-04-06 | 2017-12-13 | イセット株式会社 | 自動通報装置 |
JP2018180715A (ja) * | 2017-04-06 | 2018-11-15 | イセット株式会社 | 自動通報装置 |
WO2020049981A1 (en) * | 2018-09-05 | 2020-03-12 | Nec Corporation | An apparatus and a method for adaptively managing event-related data in a control room |
JP2021534503A (ja) * | 2018-09-05 | 2021-12-09 | 日本電気株式会社 | 制御室内のイベント関連データを適応的に管理する装置及び方法 |
JP7279776B2 (ja) | 2018-09-05 | 2023-05-23 | 日本電気株式会社 | イベント管理装置及び方法 |
Also Published As
Publication number | Publication date |
---|---|
JPWO2015198767A1 (ja) | 2017-04-20 |
US20170220871A1 (en) | 2017-08-03 |
US20200257905A1 (en) | 2020-08-13 |
JP2020098606A (ja) | 2020-06-25 |
JP7173196B2 (ja) | 2022-11-16 |
US20190205657A1 (en) | 2019-07-04 |
US11250268B2 (en) | 2022-02-15 |
JP2021106019A (ja) | 2021-07-26 |
US20230410513A1 (en) | 2023-12-21 |
US11106918B2 (en) | 2021-08-31 |
US10846536B2 (en) | 2020-11-24 |
JP7396436B2 (ja) | 2023-12-12 |
JP2023022015A (ja) | 2023-02-14 |
US20190205658A1 (en) | 2019-07-04 |
JP6696422B2 (ja) | 2020-05-20 |
JP6863488B2 (ja) | 2021-04-21 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
JP6863488B2 (ja) | 画像処理装置、異常検知方法、及びプログラム | |
WO2016157327A1 (ja) | 映像監視システム及び映像監視方法 | |
JP6504364B2 (ja) | モニタリング装置、モニタリングシステムおよびモニタリング方法 | |
US20160171332A1 (en) | Image processing apparatus, image processing method, and storage medium | |
US20120321138A1 (en) | Suspicious behavior detection system and method | |
WO2016013719A1 (ko) | 3차원 영상 정보를 이용한 이상 행동 감시 장치 및 방법 | |
JP2016048834A (ja) | モニタリング装置、モニタリングシステムおよびモニタリング方法 | |
JP6693509B2 (ja) | 画像処理装置、画像処理方法、および画像処理システム | |
JP2017028364A (ja) | 監視システムおよび監視装置 | |
JP2010063033A (ja) | 物体抽出装置、物体抽出方法および物体抽出プログラム | |
KR20200014694A (ko) | 화상 처리 장치, 화상 처리 장치의 제어 방법, 및 비일시적 컴퓨터 판독가능 저장 매체 | |
CN106210699B (zh) | 信息处理装置、信息处理装置的控制方法及图像处理*** | |
JP2023126352A (ja) | プログラム、映像監視方法及び映像監視システム | |
JP5088463B2 (ja) | 監視システム | |
JP2009147607A (ja) | 人物追尾装置 | |
US20200043177A1 (en) | Image processing device, stationary object tracking system, image processing method, and recording medium | |
JP4954459B2 (ja) | 不審人物検出装置 | |
JP2011227677A (ja) | 通報装置 | |
US10482641B2 (en) | Virtual reality display | |
JP7007323B2 (ja) | 監視表示装置 | |
JP2010063621A (ja) | 視認情景に対する視認者感性反応装置 | |
JP2020102676A (ja) | 情報処理装置、情報処理方法およびプログラム | |
JP2010171705A (ja) | 監視映像検索装置及び監視映像検索プログラム | |
KR20150124829A (ko) | 열화상 카메라를 이용한 모션 추적에 따른 실신 상황 감지 방법 및 장치 | |
JP2015186248A (ja) | モニタリング装置及びモニタリング装置の制御方法 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 15812817 Country of ref document: EP Kind code of ref document: A1 |
|
ENP | Entry into the national phase |
Ref document number: 2016529191 Country of ref document: JP Kind code of ref document: A |
|
WWE | Wipo information: entry into national phase |
Ref document number: 15321640 Country of ref document: US |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 15812817 Country of ref document: EP Kind code of ref document: A1 |