CN112850436A - Pedestrian trend detection method and system of elevator intelligent light curtain - Google Patents
Pedestrian trend detection method and system of elevator intelligent light curtain Download PDFInfo
- Publication number
- CN112850436A CN112850436A CN201911195084.8A CN201911195084A CN112850436A CN 112850436 A CN112850436 A CN 112850436A CN 201911195084 A CN201911195084 A CN 201911195084A CN 112850436 A CN112850436 A CN 112850436A
- Authority
- CN
- China
- Prior art keywords
- hall door
- detection
- frame
- door image
- current frame
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B66—HOISTING; LIFTING; HAULING
- B66B—ELEVATORS; ESCALATORS OR MOVING WALKWAYS
- B66B13/00—Doors, gates, or other apparatus controlling access to, or exit from, cages or lift well landings
- B66B13/24—Safety devices in passenger lifts, not otherwise provided for, for preventing trapping of passengers
- B66B13/26—Safety devices in passenger lifts, not otherwise provided for, for preventing trapping of passengers between closing doors
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B66—HOISTING; LIFTING; HAULING
- B66B—ELEVATORS; ESCALATORS OR MOVING WALKWAYS
- B66B5/00—Applications of checking, fault-correcting, or safety devices in elevators
- B66B5/0006—Monitoring devices or performance analysers
- B66B5/0012—Devices monitoring the users of the elevator system
Landscapes
- Indicating And Signalling Devices For Elevators (AREA)
- Elevator Door Apparatuses (AREA)
Abstract
The invention discloses a pedestrian trend detection method and a pedestrian trend detection system of an elevator intelligent light curtain, which have the technical scheme main points that the pedestrian trend detection method comprises the following steps: acquiring video data of a hoistway door gate; performing target detection on each frame of hall door image in the video data, determining the position of a target object in each frame of hall door image and marking a detection frame of the target object; based on the current frame hall door image and the nearest previous nth frame hall door image of the current frame hall door image, comparing the current frame hall door image with the position of the detection frame in the previous nth frame hall door image, and judging the motion trend of the detection frame in the current frame hall door image; and controlling the elevator to open or stop closing the door when the detection frame in the current frame hall door image has a trend of moving to one side of the elevator hall door. The invention has the characteristic of high detection accuracy.
Description
Technical Field
The invention relates to the technical field of elevator light curtains, in particular to a pedestrian trend detection method and system of an elevator intelligent light curtain.
Background
The elevator light curtain is a safety protection device for preventing passengers or goods from being clamped when an elevator is automatically closed.
At present, most of elevator light curtains are plane detection light curtains, and human bodies or objects in the motion plane of an elevator door are detected through an infrared light curtain transmitting end and an infrared light curtain receiving end which are respectively arranged at a left car door and a right car door of the elevator. By the detection mode, people or objects can be detected and the elevator can be controlled to open the door again only when the people or the objects are present in the moving plane of the elevator door and block infrared rays. However, when the elevator door is closed, passengers rush into the elevator door, and the passengers still collide with the elevator door easily to cause safety accidents.
Aiming at the defects of the common elevator light curtain, the solution commonly adopted in the industry at present is to additionally arrange a passive human body infrared detector above an elevator car door. When the elevator is ready to be closed or is closing the door, once a human body enters the detection range of the human body infrared detector, the human body infrared detector can detect the human body infrared detector by the Mashan mountain, so that the elevator immediately stops closing the door and turns to open the door. Compared with a pure plane light curtain, the door opening action is advanced in time, so that the probability of the passenger colliding with the elevator door is greatly reduced. However, this solution has the disadvantage that if the passenger wants to take the elevator in the opposite direction to the direction in which the elevator is running, the passenger just stands at the entrance of the elevator hall to wait for the change of the elevator running direction but does not want to enter the elevator, and in this case, the above detection method still opens the elevator door, so that the detection accuracy is low, and therefore there is a certain improvement.
Disclosure of Invention
Aiming at the defects in the prior art, the invention aims to provide a pedestrian trend detection method of an intelligent light curtain of an elevator, which has the characteristic of high detection accuracy.
The technical purpose of the invention is realized by the following technical scheme.
A pedestrian trend detection method of an intelligent light curtain of an elevator comprises the following steps:
acquiring video data of a hoistway door gate;
performing target detection on each frame of hall door image in the video data, determining the position of a target object in each frame of hall door image and marking a detection frame of the target object;
based on the current frame hall door image and the nearest previous nth frame hall door image of the current frame hall door image, comparing the current frame hall door image with the position of the detection frame in the previous nth frame hall door image, and judging the motion trend of the detection frame in the current frame hall door image;
and controlling the elevator to open or stop closing the door when the detection frame in the current frame hall door image has a trend of moving to one side of the elevator hall door.
Preferably, the method for judging the motion trend of the detection frame in the current frame hall door image comprises the following steps of comparing the current frame hall door image with the detection frame position in the nth frame hall door image based on the current frame hall door image and the nth frame hall door image closest to the current frame hall door image, and judging the motion trend of the detection frame in the current frame hall door image:
configuring an x-y coordinate system on each frame of hall door image based on the image acquisition range of the video data;
defining the center point of the detection frame in the current frame hall door image and the center point of the detection frame in the previous nth frame hall door image as detection points based on the current frame hall door image and the nearest previous nth frame hall door image of the current frame hall door image, and acquiring the coordinate points of the detection points;
comparing the coordinate point position of the detection point in the current frame hall door image with the coordinate point position of the detection point in the nth frame hall door image to obtain the movement speed and the movement angle of the detection point in the current frame hall door image relative to the detection point in the nth frame hall door image;
and judging the motion trend of the detection frame in the current frame hall door image according to the motion speed and the motion angle.
Preferably, the upper left corner or the upper right corner of the image capturing range of the video data is configured as the origin of the x-y coordinate system.
Preferably, when the motion angle is smaller than a preset angle and the motion speed is greater than a preset speed, it is defined that the detection frame in the current frame hall door image has a trend of moving towards one side of the elevator hall door.
Preferably, after performing target detection on each frame of hall door image in the video data, determining the position of the target object in each frame of hall door image and marking the detection frame of the target object, the method further includes the following steps:
determining the position of a target object in the current frame hall door image, marking a detection frame of the target object, and judging whether the number of the detection frames in the current frame hall door image changes relative to the number of the detection frames in the previous frame hall door image;
if yes, the current frame hall door image does not carry out detection frame motion trend judgment;
and if not, judging the motion trend of the detection frame by the current frame hall door image.
Preferably, when there is no change in the number of detection frames in the current frame hall door image relative to the number of detection frames in the previous frame hall door image, the method further includes the following steps:
the detection frame of the current frame hall door image and the detection frame of the previous frame hall door image are subjected to intersection comparison judgment;
when the intersection ratio is smaller than the preset intersection ratio, the current frame hall door image does not carry out detection frame motion trend judgment;
and when the intersection ratio is not less than the preset intersection ratio, judging the motion trend of the detection frame of the current frame hall door image.
Preferably, when the number of the detection frames in the current frame hall door image is at least two, the midpoint of the connecting line between the center points of all the detection frames in the current frame hall door image is defined as the detection point.
In view of the defects in the prior art, another object of the present invention is to provide a pedestrian trend detection system for an intelligent light curtain of an elevator, which has the characteristic of high detection accuracy.
The technical purpose of the invention is realized by the following technical scheme.
A pedestrian trend detection system of an elevator intelligent light curtain, comprising:
the acquisition unit is used for acquiring video data of a door of the elevator hall;
the detection unit is used for carrying out target detection on each frame of hall door image in the video data, determining the position of a target object in each frame of hall door image and marking out a detection frame of the target object;
the motion trend judging unit compares the current frame hall door image with the position of the detection frame in the previous nth frame hall door image based on the current frame hall door image and the nearest previous nth frame hall door image of the current frame hall door image, and judges the motion trend of the detection frame in the current frame hall door image;
and the control unit is used for controlling the elevator to open or stop closing the door when the detection frame in the current frame hall door image has the trend of moving to one side of the elevator hall door.
In summary, compared with the prior art, the beneficial effects of the invention are as follows:
according to the method, the video data of the elevator hall door gate are obtained to form a detection area at the elevator hall door gate, and the elevator door is controlled to stop closing or is reversely opened when people or pets and the like have a motion trend close to one side of the elevator hall door in a mode of carrying out target detection and judging the motion trend on each frame of hall door image in the video data; wherein, there is one or more people standing at the hoistway door gate, or when the pedestrian passes through from the hoistway door gate parallel, all can not influence the elevator and normally close the door, the condition of erroneous judgement can not appear, it is high to survey the accuracy.
Drawings
Fig. 1 is a schematic structural view of an elevator in the technical scheme of the invention;
fig. 2 is a side view of an elevator according to the solution of the invention;
fig. 3 is a top view of an elevator according to the solution of the invention;
FIG. 4 is a schematic flow chart of a pedestrian tendency detection method according to the embodiment of the present invention;
fig. 5 is a schematic flow chart of motion trend determination in the technical solution of the present invention;
FIG. 6 is a schematic diagram of the calculation of the movement speed and the movement angle in the technical solution of the present invention;
FIG. 7 is a schematic diagram of the calculation of the movement speed and the movement angle in the technical solution of the present invention;
fig. 8 is a schematic diagram of a situation two of a pedestrian in the technical solution of the present invention;
fig. 9 is a schematic diagram of a situation of a pedestrian appearing in the technical solution of the present invention;
FIG. 10 is a schematic diagram illustrating a status of cross-comparison determination of detection frames according to the present invention;
fig. 11 is a schematic diagram of a situation three of a pedestrian in the technical solution of the present invention;
fig. 12 is a schematic diagram of a situation of occurrence of a pedestrian in the technical solution of the present invention;
fig. 13 is a schematic diagram of a situation of occurrence of a pedestrian in the technical solution of the present invention;
fig. 14 is a block diagram of a pedestrian tendency detection system according to the embodiment of the present invention.
Reference numerals: 100. a camera; 200. a core control module; 300. an elevator car; 400. an elevator car door; 500. hoistway doors; 600. a wall body; 700. an infrared light curtain emission end; 800. and an infrared light curtain receiving end.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. It should be noted that: the relative arrangement of parts and steps set forth in these embodiments does not limit the scope of the present invention unless specifically stated otherwise.
The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the invention, its application, or uses. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention. Techniques, methods, and apparatus known to those of ordinary skill in the relevant art may not be discussed in detail but are intended to be part of the specification where appropriate.
Aiming at the defects of the common elevator light curtain, the commonly adopted solution in the industry at present is to install a passive human body infrared detector above the elevator car door 400, when the elevator is ready to be closed or is closing the door, once a human body enters the detection range of the human body infrared detector, the human body infrared detector can detect the human body infrared detector immediately, so that the elevator stops closing the door and opens the door immediately, but if the direction that a passenger wants to take the elevator is just opposite to the direction that the elevator is running, the passenger just stands at 500 elevator hall doors to wait for the change of the running direction of the elevator but does not want to enter the elevator, and under the condition, the detection mode still opens the elevator doors, so that the detection accuracy is lower.
Therefore, the elevator safety equipment provided by the application can be used as an elevator intelligent light curtain to solve the problem well, and by acquiring video data of 500 gates of the elevator hall door, target detection is carried out on each frame of hall door image in the video data to detect the motion trend of pedestrians, so that the elevator is controlled to open or stop closing the door, the misjudgment rate is low, and the detection accuracy is better.
The invention provides a pedestrian trend detection method of an elevator intelligent light curtain, which is applied to a core control module 200 and comprises a camera 100 and the core control module 200. Referring to fig. 3, 300 is an elevator car, 400 is an elevator car door, 500 is an elevator hall door, 600 is a wall body, and 100 is a camera, specifically, in combination with fig. 1 and 2, the elevator door includes the elevator hall door 500 and the elevator car door 400, the camera 100 is disposed on the elevator car 300 and above the elevator car door 400, and the camera 100 is used for collecting video data of the doorway of the elevator hall door 500, so that a pre-door trend protection area is formed at the doorway of the elevator hall door 500. The core control module 200 is arranged at the top of the elevator car 300, the core control module 200 is connected to an elevator control system, the core control module 200 can output a control signal to the elevator control system, and the elevator control system controls the elevator door through an elevator door motor to realize the actions of closing, opening and closing or stopping the door of the elevator door.
It is worth explaining that the elevator door can be controlled to open or stop closing the door when passengers rush in the closing process of the elevator door through the arrangement of the pre-protection area of the front trend of the elevator door, so that the elevator door is a technical scheme established on a plane detection light curtain of the elevator door. In one embodiment, an infrared light curtain transmitting end 700 and an infrared light curtain receiving end 800 are disposed between the elevator doors to form a plane detection light curtain, so as to detect human bodies or objects in the motion plane of the elevator doors to prevent the elevator doors from being pinched. In yet another embodiment, the camera 100 collects video data between the elevator doors, performs image processing and analysis on the video data to determine whether there is a foreign object, and detects whether a human body or an object in the motion plane of the elevator doors reaches the anti-pinch function of the elevator doors. The present application is not particularly limited, and the setting will be made according to actual conditions.
In view of the above technical content, if a pedestrian or an object enters the pre-protection area with a door front trend and the pedestrian or the object has a movement trend toward one side of the hoistway door 500, the core control module 200 immediately outputs a control signal to the elevator control system, and the elevator door motor is controlled by the elevator control system to control the elevator door to stop closing the door or reversely opening the door until the pedestrian or the object enters the elevator car 300 or the pedestrian or the object does not have a movement trend toward one side of the hoistway door 500, the elevator door can normally close the door.
Referring to fig. 4, the pedestrian tendency detection method of the intelligent light curtain of the elevator provided by the invention comprises the following steps:
step S100, video data of a hoistway door 500 portal is obtained;
step S200, carrying out target detection on each frame of hall door image in the video data, determining the position of the target object in each frame of hall door image and marking out the detection frame of the target object.
According to the technical scheme defined in steps 100 to 200, specifically, the camera 100 collects video data of the doorway of the hoistway door 500, and the video data is processed through the video data acquired by the camera 100. The video data consists essentially of one frame of continuous video, and in one embodiment, there are 25 hall door images per second of video data, where the time interval between each frame of hall door images is 0.04 seconds.
To determine the movement trend of the pedestrian in the video data, the position of the pedestrian in the video data needs to be identified first. According to the method, the destination detection is carried out on each frame of hall door image in the video data through an SSD destination detection algorithm so as to determine the position of a destination object in each frame of hall door image. Currently, the mainstream methods for target detection are mainly divided into two types: 1. the main idea of the two-step method (two-stage), such as R-CNN algorithm (target detection method based on convolutional neural network), is to generate a series of sparse candidate frames by heuristic method (selective search) or CNN network (convolutional neural network), and then classify and regress the candidate frames. 2. One-stage (e.g., Yolo and ssd (single Shot multi box detector)) is mainly based on the idea of uniformly and densely sampling at different positions of an image, wherein the sampling can be performed with different scales and aspect ratios, and then classification and regression are directly performed after features are extracted by using CNN. In the present application, the target detection algorithm is set according to the actual situation, and is not specifically limited.
In this embodiment, the SSD target detection algorithm is used to quickly detect the approximate location of the target object from each frame of hall door image, and then input the image information of the approximate location into a pre-trained target detection model, and the target detection model outputs the approximate location of the target object and marks out the detection frame of the target object. In one embodiment, the target detection model may employ a convolutional neural network.
In order to improve the accuracy of the target object position obtained by the output of the target detection model, the target detection model can be trained in advance, and the training process is as follows:
after a large number of sample hall door images are collected in advance, target detection can be performed on a target object in the sample hall door images through an SSD target detection algorithm to determine the approximate area position of the target object, after the approximate area position of the target object is determined, image information including the position of the target object is extracted from the sample hall door images, manual marking is further performed, label data of the sample hall door images are generated, and the label data can include all vertex coordinates of the target object and the connection sequence between the vertex coordinates, namely the detection frame of the target object.
In the training process, the collected image information of the approximate region position of the target object in the sample hall door image can be used as the input data of the target detection model, and after the input data is input into the target detection model, a corresponding output result can be obtained, wherein the output result is a detection frame of the precise region position of the target object, and the output result can be, for example, coordinate values of four vertexes arranged in sequence. Then, the output result can be compared with the label data of the sample hall door image (the label data includes the detection frame of the real position of the target object in the sample hall door image) by using a training method of the target detection model to update the model parameters of the target detection model, after a large number of sample hall door images are trained, the model parameters of the target detection model are continuously updated, and finally, the output result of the target detection model is infinitely close to the label data in the sample hall door image, so that the higher accuracy of the output result of the target detection model is ensured. Through the method, after the video data of the elevator hall door 500 door is acquired, the position of the target object in each frame of hall door image can be accurately determined, and the target object can be marked through the detection frame.
Step S300, based on the current frame hall door image and the nearest Nth frame hall door image of the current frame hall door image, comparing the positions of the detection frames in the current frame hall door image and the Nth frame hall door image, and judging the motion trend of the detection frames in the current frame hall door image;
and step S400, controlling the elevator to open or stop closing the door when the detection frame in the current frame hall door image has the trend of moving towards the elevator hall door 500 side.
According to the technical solutions defined in steps S300 to S400, specifically, after the video data is acquired, each frame of hall door image in the video data needs to be determined, in this embodiment, a current frame of hall door image is taken as an example for description, an nth frame of hall door image closest to the current frame of hall door image is taken as a reference image for motion trend determination, and the nth frame of hall door image is a third frame of hall door image in this embodiment.
In comparing the current-frame hall door image with the detection frame position in the nth-frame hall door image based on the current-frame hall door image and the nearest nth-frame hall door image, and determining the motion trend of the detection frame in the current-frame hall door image, as shown in fig. 5, the method includes the following steps:
step S310, configuring an x-y coordinate system on each frame of hall door image based on the image acquisition range of the video data;
step S320, defining the center point of the detection frame in the current frame hall door image and the center point of the detection frame in the previous frame hall door image as detection points based on the current frame hall door image and the nearest previous frame hall door image of the current frame hall door image, and acquiring the coordinate points of the detection points;
step S330, comparing the coordinate point position of the detection point in the hall door image of the current frame with the coordinate point position of the detection point in the hall door image of the previous Nth frame, and acquiring the movement speed and the movement angle of the detection point in the hall door image of the current frame relative to the detection point in the hall door image of the previous Nth frame;
and step S340, judging the motion trend of the detection frame in the current frame hall door image according to the motion speed and the motion angle.
According to the technical solutions defined in the above steps S310 to S340, specifically, the camera 100 is disposed above the elevator car door 400, and a door front trend pre-protection area is formed at the doorway of the elevator hall door 500, so that it is necessary to determine whether the pedestrian moves towards one side of the elevator hall door 500. Therefore, the upper left corner or the upper right corner of the image capturing range of the video data is configured as the origin of the x-y coordinate system, and the embodiment preferably configures the upper left corner of the image capturing range of the video data as the origin of the x-y coordinate system.
Therefore, after defining the center point of the detection frame in the current frame hall door image and the center point of the detection frame in the nth frame hall door image as the detection points, the detection points all have corresponding coordinate point positions, and the detection points fall in the first quadrant of the x-y coordinate system, namely the coordinate point positions all take positive numbers. For example, if the coordinate points of the detection frames in the current hall door image are (X1, Y1), and the coordinate points of the detection frames in the third previous hall door image are (X2, Y2), the coordinate points of the detection points in the current hall door image are compared with the coordinate points of the detection points in the third previous hall door image, and the movement speed and the movement angle of the detection points in the current hall door image relative to the detection points in the third previous hall door image are obtained.
Referring to fig. 6, the moving speed is obtained as follows, and the video data has 25 hall door images per second, so that each hall door image has a fixed time interval therebetween, and the time interval between the current hall door image and the previous third hall door image is T. Based on the coordinate points (X1, Y1) of the detection points in the hall door image of the current frame and the coordinate points (X2, Y2) of the detection points in the hall door image of the previous third frame, the displacement distance S = Y1-Y2 of the detection points in the hall door image of the current frame relative to the detection points in the hall door image of the previous third frame is calculated, so that the movement speed V = S/T = (Y1-Y2)/T of the detection points in the hall door image of the current frame relative to the detection points in the hall door image of the previous third frame. On the contrary, when the displacement distance S = Y1-Y2 is a negative number, it represents that the detection frame in the current frame hall door image has a tendency to move to the side away from the hall door 500, and the current frame hall door image does not undergo the subsequent determination process.
The process of obtaining the motion angle is as follows, since the coordinate points (X1, Y1) of the detection points in the current frame hall door image and the coordinate points (X2, Y2) of the detection points in the previous third frame hall door image are known, using the formula: arctan (| (X1-X2)/(Y1-Y2) |)/pi) 180/pi, which obtains the motion angle of the detection point in the current frame hall door image relative to the detection point in the previous third frame hall door image.
And judging the motion trend of the detection frame in the current frame hall door image according to the motion speed and the motion angle, and controlling the elevator to open or stop closing the door when the detection frame in the current frame hall door image has the motion trend towards one side of the elevator hall door 500. In the present application, when the motion angle is smaller than the preset angle and the motion speed is greater than the preset speed, it is defined that the detection frame in the current frame hall door image has a tendency to move toward the elevator hall door 500 side, and in one embodiment, the preset angle is 70 degrees and the preset speed is 0.8 m/s.
Through the mode of motion angle and motion speed dual limit, can discern that the pedestrian only stands but the health is in the swing, the pedestrian is in slight movement or the pedestrian transversely passes through circumstances such as hoistway door 500, and then can reduce the erroneous judgement rate, improve and survey the degree of accuracy.
It should be noted that, as shown in fig. 7, when the number of the detection frames in the current hall door image is at least two, the midpoint of the connecting line between the center points of all the detection frames in the current hall door image is defined as the detection point. Because the width of the hoistway door 500 is limited, the number of pedestrians standing completely in the pre-protection area for the trend in front of the door is generally two, three or four, and therefore, the number of detection frames for determining and marking the target object in each hoistway door image is generally two, three or four. In this embodiment, if the number of detection frames in the current hall door image is two, the detection frames in the current hall door image are respectively defined as a first detection frame and a second detection frame, the coordinate points (X12, Y12) of the center point of the first detection frame, the coordinate points (X22, Y22) of the center point of the second detection frame, the midpoint of the connecting line between the center points of the first detection frame and the second detection frame is a detection point, and the coordinate points ((X11 + X22)/2, (Y11+ Y22)/2) of the detection point. Similarly, the coordinate points of the center points of the first detection frame (X1, Y1), the coordinate points of the center points of the second detection frame (X2, Y2) in the third previous hall door image, the midpoint of the line connecting the center points of the first detection frame and the second detection frame is the detection point, and the coordinate points of the detection point ((X1 + X2)/2, (Y1+ Y2)/2) are determined.
Thus, in the movement speed acquisition, the displacement distance S = (Y12+ Y22)/2- (Y1+ Y2)/2 of the detection point in the current frame hall door image from the detection point in the previous third frame hall door image is calculated, and therefore, the movement speed V = S/T = ((Y12+ Y22)/2- (Y1+ Y2)/2)/T of the detection point in the current frame hall door image from the detection point in the previous third frame hall door image. On the contrary, when the displacement distance S is a negative number, it represents that the detection frame in the current frame hall door image has a movement trend toward the side away from the hall door 500, and the current frame hall door image does not perform the subsequent movement trend determination process.
In the motion angle acquisition, the formula is used: arctan (| ((X12+ X22)/2- (X1+ X2)/2)/((Y1 + Y2)/2- (Y12+ Y22)/2) |) > 180/pi, which results in the angle of motion of the detected point in the current frame hall image relative to the detected point in the previous third frame hall image.
And judging the motion trend of the detection frame in the current frame hall door image according to the motion speed and the motion angle. And when the detection frame in the current frame hall door image has a trend of moving towards one side of the elevator hall door 500, controlling the elevator to open or stop closing the door.
In order to further improve the accuracy of determining the motion trend of the detection frame in the current frame hall door image, in step 200, target detection is performed on each frame of hall door image in the video data, the position of the target object in each frame of hall door image is determined, and after the detection frame of the target object is marked, and in step 300, based on the current frame of hall door image and the N-th frame of hall door image closest to the current frame of hall door image, the current frame of hall door image is compared with the position of the detection frame in the N-th frame of hall door image, and before the motion trend of the detection frame in the current frame of hall door image is determined, the method further comprises the following steps:
the position of the target object in the current frame hall door image is determined, the detection frame of the target object is marked, and whether the number of the detection frames in the current frame hall door image changes relative to the number of the detection frames in the previous frame hall door image is judged.
When the number of the detection frames in the current frame hall door image changes relative to the number of the detection frames in the previous frame hall door image, in one embodiment, the current frame hall door image does not perform detection frame motion trend judgment to directly perform detection frame motion trend judgment in the next frame hall door image, so that the detection frames can jump in the current frame hall door image when pedestrians suddenly appear in the pre-protection area for the trend in front of the door, but the pedestrians only stand in the pre-protection area for the trend in front of the door, and the misjudgment on the current frame hall door image is avoided by skipping the motion trend judgment of the current frame hall door image.
When the number of detection frames in the current frame hall door image changes relative to the number of detection frames in the previous frame hall door image, in another embodiment, the coordinate point of the lower boundary of the current frame hall door image is used as a reference coordinate, and the reference coordinate and the coordinate point of the detection point in the current frame hall door image are subjected to calculation of the movement speed and the movement angle, so that the calculation result is ensured not to meet the condition that the detection frame in the current frame hall door image has the movement trend towards the elevator hall door 500 side. Therefore, when a pedestrian suddenly appears in the pre-protection area for the trend in front of the door, the detection frame appears in the current hall door image in a jumping mode, but the pedestrian only stands in the pre-protection area for the trend in front of the door, and misjudgment on the current hall door image is avoided by skipping the motion trend judgment of the current hall door image.
Otherwise, when the number of the detection frames in the current frame hall door image is not changed relative to the number of the detection frames in the previous frame hall door image, the current frame hall door image is used for judging the motion trend of the detection frames. It should be noted that, when the number of detection frames in the current hall door image does not change relative to the number of detection frames in the previous hall door image, the intersection and comparison determination of the detection frames is also performed, and the intersection and comparison determination further includes the following steps:
the detection frame of the current frame hall door image and the detection frame of the previous frame hall door image are subjected to intersection comparison judgment;
when the intersection ratio is smaller than the preset intersection ratio, the current frame hall door image does not carry out detection frame motion trend judgment; and otherwise, when the intersection ratio is not less than the preset intersection ratio, the current frame hall door image is used for judging the motion trend of the detection frame. In this embodiment, the predetermined intersection ratio is set to 0.5.
According to the technical scheme defined by the steps, as shown in fig. 9 and fig. 10, a situation that a pedestrian a exists in the door front trend pre-protection area at the doorway of the elevator hall door 500, the pedestrian a leaves the door front trend pre-protection area, the pedestrian b enters the door front trend pre-protection area, the pedestrian a leaves the door front trend pre-protection area, and the pedestrian b enters the door front trend pre-protection area, but the pedestrian b only passes through the door front trend pre-protection area, at this time, the position of the detection frame in the current hall door image is moved from the position of the pedestrian a to the position of the pedestrian b, a movement trend towards one side of the elevator hall door 500 direction is generated, but the number of the detection frames is not changed, so that the detection frame intersection and comparison judgment can be effectively used to avoid misjudgment.
The pedestrian trend detection method of the intelligent light curtain of the elevator can effectively cause misjudgment, improves detection accuracy and specifically comprises the following scenes:
when the situation two shown in fig. 8 occurs and a pedestrian a suddenly appears in the pre-protection area with the door front trend, the detection frame of the target pedestrian a is determined and marked in the hall door image of the current frame, and the number of the detection frames in the hall door image of the current frame changes relative to the number of the detection frames in the hall door image of the previous frame. In the present embodiment, the number of pedestrians is not particularly limited.
And in the same way, the pedestrian A exists in the pre-door trend protection area, then the pedestrian A leaves the pre-door trend protection area, and the number of the detection frames in the current frame hall door image is changed relative to the number of the detection frames in the previous frame hall door image, so that the motion trend of the current frame hall door image is not judged.
When the situation three shown in fig. 11 occurs, the pedestrian a stands on the upper side of the pre-door trend protection area, the pedestrian b enters the pre-door trend protection area from the left side, the detection frames of the pedestrian a and the pedestrian b are determined and marked in the hall door image of the current frame, which indicates that the number of the detection frames in the hall door image of the current frame changes relative to the number of the detection frames in the hall door image of the previous frame, and therefore, the hall door image of the current frame is not subjected to motion trend judgment.
If the pedestrian B crosses the pre-protection area of the pre-door trend and causes the blocking of the pedestrian A, only determining and marking a detection frame of the pedestrian B in the current frame hall door image, wherein at the moment, the number of the detection frames in the current frame hall door image is not changed relative to the number of the detection frames in the previous frame hall door image, so that the detection frames of the current frame hall door image and the detection frames of the previous frame hall door image need to be subjected to cross-over comparison judgment, and when the cross-over ratio is smaller than the preset cross-over ratio, the current frame hall door image is not subjected to detection frame motion trend judgment; and otherwise, when the intersection ratio is not less than the preset intersection ratio, the current frame hall door image is used for judging the motion trend of the detection frame.
When the situation four shown in fig. 12 occurs, the pedestrian a is at the upper half part of the pre-protection area for the trend in front of the door, the pedestrian b is at the lower half part of the pre-protection area for the trend in front of the door, and the pedestrian a and the pedestrian b move away from one side of the elevator hall door 500 at the same time, when the pedestrian a is located at the boundary of the pre-protection area for the trend in front of the door, the detection frame of the pedestrian a cannot be determined and marked in the hall door image of the current frame, so that the detection frame jumps from the position of the pedestrian a to the position of the pedestrian b in the hall door image of the current frame, and a motion trend toward one side of the elevator hall door 500 is generated, so that in order to avoid misjudgment, the number of the detection frames in the hall door image changes relative to the number of the detection frames in the hall door image of the previous frame, and therefore, the motion trend.
If the pedestrian B needs to temporarily return to the elevator car 300 at the moment, the pedestrian B is very close to the elevator hall door 500, and the door opening or the door closing of the elevator can be controlled by directly separating the elevator hall door 500 or the elevator car door 400 by hands.
However, only when the situation five shown in fig. 13 occurs, for example, the pedestrian a is in the upper half of the pre-protection area for the trend in front of the door, the pedestrian b is in the lower half of the pre-protection area for the trend in front of the door, the pedestrian a moves towards the side of the elevator hall door 500, and the pedestrian b moves away from the side of the elevator hall door 500, wherein the detection frames of the pedestrian a and the pedestrian b are determined and marked in the image of the elevator hall door 500 at the current frame, but the movement speed of the pedestrian a is less than that of the pedestrian b, and at this time, a misjudgment situation may occur. However, as proved by a large number of practical tests, the situation that the pedestrian A collides with the elevator hall door 500 cannot occur during the normal door closing period of the elevator due to the fact that the movement speed of the pedestrian A is low.
Through the explanation of the situation, after the video data is obtained, the method can perform target detection on each frame of hall door image in the video data, determine the position of the target object in each frame of hall door image and mark the detection frame of the target object, judge whether the number of the detection frames in the current frame of hall door image changes relative to the number of the detection frames in the previous frame of hall door image in advance before judging the motion trend of the detection frames in the current frame of hall door image, namely, do not perform motion trend judgment on the current frame of hall door image relative to the number of the detection frames in the previous frame of hall door image, and directly perform motion trend judgment on the next frame of hall door image so as to avoid the occurrence of misjudgment;
and when the number of the detection frames in the current frame hall door image is not changed relative to the number of the detection frames in the previous frame hall door image, further comparing the intersection ratio of the detection frames in the current frame hall door image and the detection frames in the previous frame hall door image, and judging the motion trend of the detection frames in the current frame hall door image only after the intersection ratio of the detection frames meets the requirement, so as to further avoid the occurrence of misjudgment and improve the detection accuracy.
The invention also provides a pedestrian trend detection system of the intelligent elevator light curtain, which is applied to a core control module 200 of elevator safety equipment, and as shown in fig. 14, the pedestrian trend detection system of the intelligent elevator light curtain comprises an acquisition unit, a detection unit, a motion trend judgment unit and a control unit.
The obtaining unit is used for obtaining video data of the hoistway door 500; the detection unit is used for carrying out target detection on each frame of hall door image in the video data, determining the position of a target object in each frame of hall door image and marking out a detection frame of the target object.
The motion trend judging unit compares the current frame hall door image with the position of the detection frame in the previous nth frame hall door image based on the current frame hall door image and the nearest previous nth frame hall door image of the current frame hall door image, and judges the motion trend of the detection frame in the current frame hall door image; the control unit is used for controlling the elevator to open or stop closing the door when the detection frame in the current frame hall door image has the trend of moving towards the elevator hall door 500 side.
The motion trend judging unit comprises an establishing unit, a defining unit, a calculating unit and a judging unit.
The establishing unit configures an x-y coordinate system on each frame of hall door image based on the image capturing range of the video data, and it should be noted that, in this embodiment, the upper left corner or the upper right corner of the image capturing range of the video data is configured as the origin of the x-y coordinate system, and is preferably the upper left corner of the image capturing range of the video data.
The defining unit defines the center point of the detection frame in the current frame hall door image and the center point of the detection frame in the previous nth frame hall door image as detection points based on the current frame hall door image and the nearest previous nth frame hall door image of the current frame hall door image, and obtains the coordinate point positions of the detection points. It should be noted that, when the number of detection frames in the current hall door image is at least two, the midpoint of the connecting line between the center points of all the detection frames in the current hall door image is defined as a detection point.
The calculating unit is used for comparing the coordinate point position of the detection point in the current frame hall door image with the coordinate point position of the detection point in the nth frame hall door image to obtain the movement speed and the movement angle of the detection point in the current frame hall door image relative to the detection point in the nth frame hall door image; the judging unit is used for judging the motion trend of the detection frame in the current frame hall door image according to the motion speed and the motion angle. When the motion angle is smaller than a preset angle and the motion speed is greater than a preset speed, it is defined that the detection frame in the current frame hall door image has a trend of moving towards one side of the elevator hall door 500.
The pedestrian trend detection system of the elevator intelligent light curtain further comprises a frame quantity detection unit, wherein the frame quantity detection unit is used for determining the position of a target object in the current frame hall door image, marking a detection frame of the target object and judging whether the quantity of the detection frames in the current frame hall door image changes relative to the quantity of the detection frames in the previous frame hall door image.
If the number of the detection frames in the current frame hall door image changes relative to the number of the detection frames in the previous frame hall door image, the current frame hall door image does not carry out detection frame motion trend judgment;
on the contrary, if the number of the detection frames in the current frame hall door image is not changed relative to the number of the detection frames in the previous frame hall door image, the current frame hall door image is used for judging the motion trend of the detection frames.
It should be noted that, in order to further improve the detection accuracy, when the number of detection frames in the current frame hall door image does not change relative to the number of detection frames in the previous frame hall door image, the intersection and comparison determination of the detection frames is also performed, and the intersection and comparison determination of the detection frames includes:
the detection frame of the current frame hall door image and the detection frame of the previous frame hall door image are subjected to intersection comparison judgment;
when the intersection ratio is smaller than the preset intersection ratio, the current frame hall door image does not carry out detection frame motion trend judgment;
and when the intersection ratio is not less than the preset intersection ratio, judging the motion trend of the detection frame of the current frame hall door image.
After video data are obtained, target detection can be carried out on each frame of hall door image in the video data, the position of a target object in each frame of hall door image is determined, a detection frame of the target object is marked, whether the number of the detection frames in the current frame of hall door image changes relative to the number of the detection frames in the previous frame of hall door image is judged in advance before the detection frames in the current frame of hall door image are subjected to motion trend judgment, the number of the detection frames in the current frame of hall door image changes relative to the number of the detection frames in the previous frame of hall door image is judged, namely, the motion trend judgment is not carried out on the current frame of hall door image, and the motion trend judgment is directly carried out on the next frame of hall door image, so that the condition of misjudgment is avoided;
and when the number of the detection frames in the current frame hall door image is not changed relative to the number of the detection frames in the previous frame hall door image, further comparing the intersection ratio of the detection frames in the current frame hall door image and the detection frames in the previous frame hall door image, and judging the motion trend of the detection frames in the current frame hall door image only after the intersection ratio of the detection frames meets the requirement, so as to further avoid the occurrence of misjudgment and improve the detection accuracy.
The above description is intended to be illustrative of the present invention and not to limit the scope of the invention, which is defined by the claims appended hereto.
Claims (8)
1. A pedestrian trend detection method of an intelligent light curtain of an elevator is characterized by comprising the following steps:
acquiring video data of a doorway of a hoistway door (500);
performing target detection on each frame of hall door image in the video data, determining the position of a target object in each frame of hall door image and marking a detection frame of the target object;
based on the current frame hall door image and the nearest previous nth frame hall door image of the current frame hall door image, comparing the current frame hall door image with the position of the detection frame in the previous nth frame hall door image, and judging the motion trend of the detection frame in the current frame hall door image;
and when the detection frame in the current frame hall door image has a trend of moving towards one side of the elevator hall door (500), controlling the elevator to open or stop closing the door.
2. The pedestrian trend detection method of the intelligent light curtain of the elevator as claimed in claim 1, wherein based on the current frame hall door image and the nearest previous nth frame hall door image of the current frame hall door image, the current frame hall door image is compared with the detection frame position in the previous nth frame hall door image, and the motion trend of the detection frame in the current frame hall door image is judged, comprising the following steps:
configuring an x-y coordinate system on each frame of hall door image based on the image acquisition range of the video data;
defining the center point of the detection frame in the current frame hall door image and the center point of the detection frame in the previous nth frame hall door image as detection points based on the current frame hall door image and the nearest previous nth frame hall door image of the current frame hall door image, and acquiring the coordinate points of the detection points;
comparing the coordinate point position of the detection point in the current frame hall door image with the coordinate point position of the detection point in the nth frame hall door image to obtain the movement speed and the movement angle of the detection point in the current frame hall door image relative to the detection point in the nth frame hall door image;
and judging the motion trend of the detection frame in the current frame hall door image according to the motion speed and the motion angle.
3. The pedestrian tendency detection method of the elevator intelligent light curtain as claimed in claim 2, wherein the upper left corner or the upper right corner of the image acquisition range of the video data is configured as the origin of an x-y coordinate system.
4. The pedestrian tendency detection method of the intelligent light curtain of the elevator as claimed in claim 3, wherein when the motion angle is smaller than a preset angle and the motion speed is greater than a preset speed, the detection frame in the current frame hall door image is defined to have a tendency to move towards the side of the elevator hall door (500).
5. The pedestrian trend detection method of the intelligent light curtain of the elevator as claimed in claim 1, wherein after the target detection is performed on each frame of hall door image in the video data, the position of the target object in each frame of hall door image is determined and the detection frame of the target object is marked, the method further comprises the following steps:
determining the position of a target object in the current frame hall door image, marking a detection frame of the target object, and judging whether the number of the detection frames in the current frame hall door image changes relative to the number of the detection frames in the previous frame hall door image;
if yes, the current frame hall door image does not carry out detection frame motion trend judgment;
and if not, judging the motion trend of the detection frame by the current frame hall door image.
6. The pedestrian tendency detection method of the intelligent light curtain of the elevator as claimed in claim 5, wherein when the number of the detection frames in the current frame hall door image is not changed relative to the number of the detection frames in the previous frame hall door image, the method further comprises the following steps:
the detection frame of the current frame hall door image and the detection frame of the previous frame hall door image are subjected to intersection comparison judgment;
when the intersection ratio is smaller than the preset intersection ratio, the current frame hall door image does not carry out detection frame motion trend judgment;
and when the intersection ratio is not less than the preset intersection ratio, judging the motion trend of the detection frame of the current frame hall door image.
7. The pedestrian tendency detection method of the intelligent light curtain of the elevator as claimed in claim 2, wherein when the number of the detection frames in the current frame hall door image is at least two, the midpoint of the connecting line between the central points of all the detection frames in the current frame hall door image is defined as the detection point.
8. The utility model provides a pedestrian trend detecting system of elevator intelligence light curtain which characterized in that includes:
an acquisition unit for acquiring video data of a doorway of a hoistway door (500);
the detection unit is used for carrying out target detection on each frame of hall door image in the video data, determining the position of a target object in each frame of hall door image and marking out a detection frame of the target object;
the motion trend judging unit compares the current frame hall door image with the position of the detection frame in the previous nth frame hall door image based on the current frame hall door image and the nearest previous nth frame hall door image of the current frame hall door image, and judges the motion trend of the detection frame in the current frame hall door image;
and the control unit is used for controlling the elevator to open or stop closing the door when the detection frame in the current frame hall door image has the trend of moving towards one side of the elevator hall door (500).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911195084.8A CN112850436A (en) | 2019-11-28 | 2019-11-28 | Pedestrian trend detection method and system of elevator intelligent light curtain |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911195084.8A CN112850436A (en) | 2019-11-28 | 2019-11-28 | Pedestrian trend detection method and system of elevator intelligent light curtain |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112850436A true CN112850436A (en) | 2021-05-28 |
Family
ID=75995940
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911195084.8A Pending CN112850436A (en) | 2019-11-28 | 2019-11-28 | Pedestrian trend detection method and system of elevator intelligent light curtain |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112850436A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115849144A (en) * | 2022-09-22 | 2023-03-28 | 杭州展特智能科技有限公司 | Cargo identification and anti-collision control method for cargo lift |
| CN116386195A (en) * | 2023-05-29 | 2023-07-04 | 南京致能电力科技有限公司 | Face access control system based on image processing |
Citations (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2004157879A (en) * | 2002-11-07 | 2004-06-03 | Matsushita Electric Works Ltd | Object detection method |
| CN101723226A (en) * | 2009-12-24 | 2010-06-09 | 杭州优迈科技有限公司 | System and method of machine vision three-dimensional detection elevator light curtain |
| CN102701056A (en) * | 2011-12-12 | 2012-10-03 | 广州都盛机电有限公司 | Dynamic image recognition control method, device and system |
| CN103413321A (en) * | 2013-07-16 | 2013-11-27 | 南京师范大学 | Crowd behavior model analysis and abnormal behavior detection method under geographical environment |
| CN106044502A (en) * | 2016-08-15 | 2016-10-26 | 广东新力欧菲尔电梯有限公司 | Elevator door anti-collision protection system and using method |
| CN106203274A (en) * | 2016-06-29 | 2016-12-07 | 长沙慧联智能科技有限公司 | Pedestrian's real-time detecting system and method in a kind of video monitoring |
| CN108222749A (en) * | 2017-12-29 | 2018-06-29 | 安徽慧视金瞳科技有限公司 | A kind of intelligent automatic door control method based on image analysis |
| CN109064484A (en) * | 2018-03-21 | 2018-12-21 | 南京邮电大学 | Crowd movement's Activity recognition method with momentum Fusion Features is divided based on subgroup |
| CN109325404A (en) * | 2018-08-07 | 2019-02-12 | 长安大学 | A kind of demographic method under public transport scene |
| CN109670503A (en) * | 2018-12-19 | 2019-04-23 | 北京旷视科技有限公司 | Label detection method, apparatus and electronic system |
| CN109850730A (en) * | 2018-11-20 | 2019-06-07 | 日立楼宇技术（广州）有限公司 | The discrepancy door control method of particular space, device and system |
| CN109886997A (en) * | 2019-01-23 | 2019-06-14 | 平安科技（深圳）有限公司 | Method, apparatus and terminal device are determined based on the identification frame of target detection |
| CN110084831A (en) * | 2019-04-23 | 2019-08-02 | 江南大学 | Based on the more Bernoulli Jacob's video multi-target detecting and tracking methods of YOLOv3 |
| CN110287950A (en) * | 2019-06-05 | 2019-09-27 | 北京字节跳动网络技术有限公司 | Target detection and the training method of target detection model, device and electronic equipment |
| CN110378420A (en) * | 2019-07-19 | 2019-10-25 | Oppo广东移动通信有限公司 | A kind of image detecting method, device and computer readable storage medium |
| CN110400332A (en) * | 2018-04-25 | 2019-11-01 | 杭州海康威视数字技术股份有限公司 | A kind of target detection tracking method, device and computer equipment |
-
2019
- 2019-11-28 CN CN201911195084.8A patent/CN112850436A/en active Pending
Patent Citations (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2004157879A (en) * | 2002-11-07 | 2004-06-03 | Matsushita Electric Works Ltd | Object detection method |
| CN101723226A (en) * | 2009-12-24 | 2010-06-09 | 杭州优迈科技有限公司 | System and method of machine vision three-dimensional detection elevator light curtain |
| CN102701056A (en) * | 2011-12-12 | 2012-10-03 | 广州都盛机电有限公司 | Dynamic image recognition control method, device and system |
| CN103413321A (en) * | 2013-07-16 | 2013-11-27 | 南京师范大学 | Crowd behavior model analysis and abnormal behavior detection method under geographical environment |
| CN106203274A (en) * | 2016-06-29 | 2016-12-07 | 长沙慧联智能科技有限公司 | Pedestrian's real-time detecting system and method in a kind of video monitoring |
| CN106044502A (en) * | 2016-08-15 | 2016-10-26 | 广东新力欧菲尔电梯有限公司 | Elevator door anti-collision protection system and using method |
| CN108222749A (en) * | 2017-12-29 | 2018-06-29 | 安徽慧视金瞳科技有限公司 | A kind of intelligent automatic door control method based on image analysis |
| CN109064484A (en) * | 2018-03-21 | 2018-12-21 | 南京邮电大学 | Crowd movement's Activity recognition method with momentum Fusion Features is divided based on subgroup |
| CN110400332A (en) * | 2018-04-25 | 2019-11-01 | 杭州海康威视数字技术股份有限公司 | A kind of target detection tracking method, device and computer equipment |
| CN109325404A (en) * | 2018-08-07 | 2019-02-12 | 长安大学 | A kind of demographic method under public transport scene |
| CN109850730A (en) * | 2018-11-20 | 2019-06-07 | 日立楼宇技术（广州）有限公司 | The discrepancy door control method of particular space, device and system |
| CN109670503A (en) * | 2018-12-19 | 2019-04-23 | 北京旷视科技有限公司 | Label detection method, apparatus and electronic system |
| CN109886997A (en) * | 2019-01-23 | 2019-06-14 | 平安科技（深圳）有限公司 | Method, apparatus and terminal device are determined based on the identification frame of target detection |
| CN110084831A (en) * | 2019-04-23 | 2019-08-02 | 江南大学 | Based on the more Bernoulli Jacob's video multi-target detecting and tracking methods of YOLOv3 |
| CN110287950A (en) * | 2019-06-05 | 2019-09-27 | 北京字节跳动网络技术有限公司 | Target detection and the training method of target detection model, device and electronic equipment |
| CN110378420A (en) * | 2019-07-19 | 2019-10-25 | Oppo广东移动通信有限公司 | A kind of image detecting method, device and computer readable storage medium |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115849144A (en) * | 2022-09-22 | 2023-03-28 | 杭州展特智能科技有限公司 | Cargo identification and anti-collision control method for cargo lift |
| CN115849144B (en) * | 2022-09-22 | 2024-04-26 | 杭州展特智能科技有限公司 | Goods identification and anti-collision control method for goods elevator |
| CN116386195A (en) * | 2023-05-29 | 2023-07-04 | 南京致能电力科技有限公司 | Face access control system based on image processing |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP2801956B1 (en) | Passenger counter | |
| US8660700B2 (en) | Video-based system and method of elevator door detection | |
| US5554983A (en) | Object recognition system and abnormality detection system using image processing | |
| US20190185295A1 (en) | User detection system and image processing device | |
| CN108280952B (en) | Passenger trailing monitoring method based on foreground object segmentation | |
| US20130070969A1 (en) | Method and system for people flow statistics | |
| CN104573659B (en) | A kind of driver based on svm takes phone-monitoring method | |
| CN105744232A (en) | Method for preventing power transmission line from being externally broken through video based on behaviour analysis technology | |
| CN112850436A (en) | Pedestrian trend detection method and system of elevator intelligent light curtain | |
| CN112669497A (en) | Pedestrian passageway perception system and method based on stereoscopic vision technology | |
| US20210225091A1 (en) | Auto-calibrating range sensing gate | |
| KR20190046351A (en) | Method and Apparatus for Detecting Intruder | |
| CN116402863B (en) | Intelligent analysis and early warning system for building construction monitoring data | |
| CN116153086B (en) | Multi-path traffic accident and congestion detection method and system based on deep learning | |
| CN110874551A (en) | Image processing method for access control gate passage logic control | |
| CN108109304B (en) | Subway platform intruder recognition system and recognition method | |
| JP2002293484A (en) | Elevator control device | |
| CN113128322A (en) | Elevator sill detection method | |
| CN111104554B (en) | Automatic door control method and system based on contour detection and head and shoulder detection | |
| CN111646330A (en) | Elevator detection device | |
| CN115258865A (en) | Identification method and device for elevator door | |
| CN112723077A (en) | Elevator door opening and closing detection method based on optical flow | |
| CN113247720A (en) | Intelligent elevator control method and system based on video | |
| CN109472218A (en) | A kind of automobile assistant driving safety method based on pedestrian's direction discernment | |
| KR101254393B1 (en) | Method and device for detecting head position of traveling body |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210528 |