CN102917171A - Small target locating method based on pixel - Google Patents

Abstract

The invention relates to the field of video monitoring, and provides a small target locating linkage monitoring method based on a pixel by linkage of a fixed camera and a rotating camera. The small target locating method comprises two steps of capturing a target and locating a target; the step of capturing the target is as follows: triggering to start the rotating camera by selecting a panoramic video of the fixed camera; the step of locating the target is as follows obtaining an intact high-definition video picture of the located target through the rotating camera, scanning image outline points, and converting video outline point information into a chart coordinate, and projecting on a chart, creating a projection position point on the chart is created into an area object through a chart area drawing tool, directly obtaining longitude and latitude information of the position on the chart by calculating a geometric center of the area, and then setting lock monitor of a video camera on the target through the obtained longitude and latitude information of the target position to realize the real-time dynamic state. According to the method, the panoramic video picture is provided to the rotating camera through the fixed camera, and the rotating camera accurately captures and locates the chosen target.

Description

Little object localization method based on pixel

Technical field

The present invention relates to field of video monitoring, by the fixedly interlock of camera and rotary camera, realize the little target localization interlock method for supervising based on pixel.

Background technology

Along with developing rapidly of computer technology and image processing techniques, for improving the safety monitoring ability to strengthen law enforcement dynamics, needs obtain clear and monitor video is as the law enforcement foundation accurately, and users at different levels improve day by day to the video monitoring demand of intellectuality, accuracy, high Qinghua and removableization.For a long time, video monitoring mainly is confined to the fixed area monitoring of fixing camera, is subjected to the impact of pixel and intelligent degree, and video monitoring has the zone limitation, definition is low and the problem such as single picture disply, brings difficulty for monitoring and law enforcement evidence obtaining.So need a kind of panorama, various visual angles monitoring and target localization of realizing the video monitoring target with the monitoring mode of being association of activity and inertia.

Summary of the invention

The invention provides a kind of little object localization method based on pixel, realize monitoring and location to different target in the same picture.

For achieving the above object, the present invention takes following technical scheme:

Little object localization method based on pixel may further comprise the steps:

(1) target acquisition

1) videoscanning: fixedly camera scans monitoring to the appointed area in the field range;

2) view image triggers sound and light alarm automatically through the backstage analytical calculation when meeting the default alarm rule;

3) according to alert locations information, carry out alarming determining by the fixing panoramic video of camera;

4) select emphasis scene and target in the video pictures by frame selection tool frame, the frame selection operation triggers the startup of rotary camera;

5) obtain frame selected scenes scape geometric center location point coordinate by hind computation, and according to matching algorithm, calculate and obtain the position of this geometric center in rotary camera, thereby obtain the cloud platform rotation angle;

6) regulate rotation camera head orientation, zoom obtains the HD video of scape position, frame selected scenes;

(2) target localization

1) regulates by the video of rotary camera, obtain the complete HD video picture of localizing objects;

2) carry out the image outline described point at the HD video picture, and by the back-end data processing center, the visual profiles dot information is converted to the sea chart coordinate, project on the sea chart;

3) the projected position point on the sea chart is created as section object by sea chart zone drawing tool, by calculating this regional geometric center, directly obtains the latitude and longitude information of this position at sea chart;

4) by the sea chart referential function, event icons and explanation are added in this position, convenient users at different levels check and share;

5) the target location latitude and longitude information by obtaining arranges the locking monitoring to this target of video camera, and is dynamic in real time to understand it;

6) send event report information to relevant administrative department.

The present invention is that rotary camera offers real-time, accurate and complete panoramic video picture by fixing camera, and in the panoramic video picture to the frame choosing of highest priority, start rotary camera and realize accurate seizure and location to highest priority; Rotary camera realizes that by rotation and the Focussing of rotary camera high definition, amplification to localized target in the panoramic video picture show then by gain of parameter localized target or regional positional informations such as camera pixel, focal length and visual fields.

Description of drawings

Fig. 1 is FB(flow block) of the present invention;

Fig. 2 is rotary camera, fixedly camera and frame select the geometrical relationship figure of target;

Fig. 3 is the schematic diagram of camera coordinate system;

Fig. 4 is the schematic diagram of image coordinate system;

Fig. 5 is the schematic diagram that pixel image catches.

Embodiment

Below in conjunction with drawings and Examples content of the present invention is described further.

Realize that capital equipment of the present invention comprises fixedly camera and rotary camera, they all are high-definition cameras, and they all can carry out the adjusting of night vision pattern and visible light pattern.Fixedly camera is the principal phase machine of video monitoring system system default, keep fixing azimuth and fixing field range, can not carry out any change of camera position and visual field, fixedly the video pictures of camera is the panoramic video that can arrive, and also has following operating function:

(1) visual field mapping: the visual field image of video camera is become corresponding geographic orientation and distance with the form responding of pixel, and determine the geographic orientation of every bit in the image with this;

(2) target frame choosing: arbitrary region in the visual field is chosen and located by selecting the scape frame;

(3) instruction generates: select the positional information of scape frame to generate corresponding orientation instruction, the driving instruction that the video camera that rotates can be identified;

(4) command: send video camera (The Cloud Terrace and motor-driven lens) instruction fixed point to corresponding rotary camera.

The rotation of rotary camera by The Cloud Terrace realizes 360 ° zone and target monitoring, and it can carry out the self-defined adjusting of position and visual field, can carry out high definition to monitoring objective by zoom and amplify and show, is additional and perfect to fixing phase machine monitoring.Also have following operating function:

(1) local operation: The Cloud Terrace and camera lens by rotary camera carry out common rotation and Focussing operation;

(2) combined operation:

1) command reception: receive the instruction of issuing this video camera;

2) target localization: according to the size in zone, calculate the best monitoring visual field of target area in conjunction with the resolution of rotating video camera, focal length etc.;

3) camera driver: actuated camera is to specific orientation and focal length.

Realize operating procedure and the function of supervisory control system of the present invention:

(1) obtains panoramic video

By video monitoring system, startup fixedly camera is fixed regional panoramic video scanning, obtains the panoramic video picture of fixed area, and the panoramic video picture can by monitoring window layout pattern, carry out the demonstration of single window, multiwindow form.

(2) selected monitoring objective

In the panoramic video picture of fixing camera, highest priority is passed through the fixedly picture zoom function of camera, carry out target investigation and examination, the user is concerned about, the not good destination object of display effect carries out monitoring objective by the frame selection tool on the panoramic video picture selectes, by cloud platform rotation and its detailed picture information of zoom amplification demonstration of rotary camera.

(3) start rotary camera

After the selected local monitoring objective of frame selection tool, drag the frame favored area to the panoramic video picture to start rotary camera, realize to rotate video camera and the fixing video interlink of camera, then discharge mouse, the high definition amplified video window of acquisition rotary camera selected target.After obtaining the HD video picture, can according to actual user's demand, carry out picture and extract the operations such as video record.

(4) rotary camera target localization

1) determines that frame selects target in fixing magazine position;

2) setting up fixedly, the three-dimensional coordinate space of camera and rotary camera relative position is;

3) by fixing camera focus f, the fixing angle of camera and horizontal plane, calculation block is selected the position vector of target on fixing camera;

4) in three-dimensional system of coordinate, calculate rotary camera with respect to the fixing anglec of rotation of camera, and the vector position of rotary camera in the XOY coordinate surface;

5) based on rotational angle and the position movement parameter of calculating gained, rotary platform, captured target.

Wherein the target sizes decision method of rotary camera is as follows

After rotary camera obtains frame and selects target, obtain the HD video picture that frame selects target by zoom, this moment by fixing camera focal distance f and the optical center to the subject distance D, calculate the size of acquisition monitoring objective, its computing formula is as follows:

H=h*D/f calculates the horizontal range of photographic subjects;

V=v*D/f calculates the vertical range of photographic subjects;

Wherein: D: the optical center of rotary camera is to subject distance (geometric center distance)

F: rotary camera focal length

H: the CCD sensitization target surface horizontal range of rotary camera

V: the CCD sensitization target surface horizontal range of rotary camera A

H: the horizontal range of subject

V: the vertical range of subject

As shown in Figure 1, the present invention realizes little object localization method based on pixel by fixedly camera and rotary camera video interlink, and to realize monitoring and the location to different target in the same picture, concrete Computer Processing flow process is as follows:

(1) target acquisition

1) videoscanning: fixedly camera scans monitoring to the appointed area in the field range;

2) view image triggers sound and light alarm automatically through the backstage analytical calculation when meeting the default alarm rule;

3) according to alert locations information, carry out alarming determining by the fixing panoramic video of camera;

4) select emphasis scene and target in the video pictures by frame selection tool frame, the frame selection operation triggers the startup of rotary camera;

5) obtain frame selected scenes scape geometric center location point coordinate by hind computation, and according to matching algorithm, calculate and obtain the position of this geometric center in rotary camera, thereby obtain the cloud platform rotation angle;

6) regulate rotation camera head orientation, zoom obtains the HD video of scape position, frame selected scenes;

(2) target localization

1) regulates by the video of rotary camera, obtain the complete HD video picture of localizing objects;

2) carry out the image outline described point at the HD video picture, and by the back-end data processing center, the visual profiles dot information is converted to the sea chart coordinate, project on the sea chart;

3) the projected position point on the sea chart is created as section object by sea chart zone drawing tool, by calculating this regional geometric center, directly obtains the latitude and longitude information of this position at sea chart;

4) by the sea chart referential function, event icons and explanation are added in this position, convenient users at different levels check and share;

5) the target location latitude and longitude information by obtaining arranges the locking monitoring to this target of video camera, and is dynamic in real time to understand it;

6) send event report information to relevant administrative department.

The target localization algorithmic procedure is as follows:

In video frequency following system, the frame selection tool is passed through in the seizure of target, destination object is extracted from background, image is carried out real-time dividing processing, thereby obtain a width of cloth bianry image.For the target area information that comprises in the detected image, needing the mark gray value is 1 pixel, then checks the connectedness between the neighborhood pixels, will have the pixel of connected relation as a target area, extracts relevant characteristic quantity.

Frame selects the seizure of target to calculate coupling by grey scale pixel value and obtains, the calculating of geometric center position coordinates is by the camera coordinate system of foundation and the Coordinate Conversion between image coordinate system, utilize the relation of camera pixel, resolution and focal length, to obtain the coordinate position of targets of interest.

Pixel coordinate Calculation: width dimensions x resolution=width pixel;

Height dimension x resolution=height pixel.

As shown in Figure 2,

A: the fixing position of camera; B: the position of rotary camera; C: frame selects the target geometric center;

A1: the fixing angle of camera and y direction is fixing known parameters;

B1: the angle of rotary camera and y direction is unknown parameter;

C1: the fixing relative position angle of camera and rotary camera is unknown parameter.

Suppose,

1) fixedly height with identical Y-direction of camera and rotary camera, i.e. AA1=BB1=L;

2) fixedly camera and rotary camera keep certain positional distance, and namely AB is fixed value;

3) fixedly camera and rotary camera frame is selected the centre of location of target C is the geometric center that frame selects grid, geometric center is obtained by fixing camera auto-focus, namely guarantee rotary camera and fixedly the imaging point of camera overlap.

Geometric algorithm is as follows:

Known parameters is as follows:

1) the fixing angle a1 of camera and y direction;

2) rotary camera has the height L of identical Y-direction, AA1=BB1=L;

d:AB=A1B1=d

In right-angled triangle AA1C: AC=AA1/cosa1;

A1C=AC*sina1=AA1*tga1；

In right-angled triangle BB1C: according to universal formula, can obtain the rotational angle of rotary camera The Cloud Terrace in the YOZ plane and be:

$\mathrm{BC}=\sqrt{d^2-{\left(\mathrm{ltga}1\right)}^2}$

$\mathrm{<figure-callout\; id="1"\; label="b"\; filenames="HDA00002287109600024.png"\; state="\{\{state\}\}">b</figure-callout>}1=\mathrm{arctg}\sqrt{d^2-{\left(\mathrm{ltga}1\right)}^2}/l;$

In triangle ABC, according to the cosine law: AB ²=AC ²+ BC ²-2*AC*BC*cosc1,

Rotary camera with respect to the fixing angle that horizontally rotates of camera is as can be known:

${\mathrm{<figure-callout\; id="2"\; label="d"\; filenames="HDA00002287109600024.png"\; state="\{\{state\}\}">d</figure-callout>}}^2={\mathrm{<figure-callout\; id="2"\; label="l"\; filenames="HDA00002287109600024.png"\; state="\{\{state\}\}">l</figure-callout>}}^2+d^2-{\left(\mathrm{ltga}1\right)}^2-2*l/\cos a1*\sqrt{d^2-{\left(\mathrm{ltga}1\right)}^2}<figure-callout\; id="1"\; label="*\; cos\; c"\; filenames="HDA00002287109600024.png"\; state="\{\{state\}\}">*</figure-callout>\cos c1$

That is: $c1=\mathrm{<figure-callout\; id="1"\; label="arccos"\; filenames="HDA00002287109600024.png"\; state="\{\{state\}\}">arccos</figure-callout>}\frac{1^2-{\left(\mathrm{ltga}1\right)}^2}{2-1/\cos a1\sqrt{d^2-{\left(\mathrm{ltga}1\right)}^2}}$

The target localization algorithm is as follows:

(1) as shown in Figure 3, set up camera coordinate system;

(2) as shown in Figure 4, set up image coordinate system: image coordinate is divided into pixel coordinate system and physical coordinates system.Wherein image coordinate system is take pixel as unit, take the image upper left corner as initial point, take u, v direction as two reference axis.Image physical coordinates system is with the focus O of optical axis with the picture plane ₁Rectangular coordinate system for initial point.

(3) as shown in Figure 5, catch algorithm examples based on pixel image as follows:

1) the 1st scan line begins a new target label S (1) in the position of pixel (1,3); In the time of processed pixels point (1,6), learn mark S (1).Finish at pixel (1,5) point, write the target label among the NS among the S (1) this moment;

2) the 2nd scan line, the position in pixel (2,2), by algorithm as can be known, it belongs to target label S (1).The 3rd scan line also can similarly be processed;

3) (4, the 6) pixel from the 4th scan line begins a new target label S (2).The pixel of the 5th scan line belongs to respectively target label S (1) and S (2).

4) the 6th scan line, from pixel (6,1) new target label S (3) of beginning, when scanning pixel (6,2) time, according to algorithm, target label S (1) will merge among the S (3), it is invalid that the significant notation of while S (1) is put, and equivalent indicia is pointed to S (3); When scanning pixel (6,6), merge the content of S (2) in S (3); When scanning pixel (6,8), because S (2) significant notation zero setting, and equivalent indicia is pointed to S (3) itself, so do not do any action; When scanning pixel (6,11), content among the NS is written to S (3), the content among S this moment (3) is S (1), the merging of S (2) and NS.

5) the 6th scan line that coexists in pixel (6,13), is opened up fresh target mark S (4); At pixel (6,14), check (5,15) point corresponding target label S (2), its significance bit is zero, S (3) is pointed in the equivalence position, so the content of S (3) is write S (4), it is invalid that S (3) is set to, equivalent indicia is pointed to S (4), check simultaneously all target labels, find original S (1) and S (2) equivalent indicia sensing S (3), existing equivalent indicia with S (1) and S (2) should be to point to S (4).In pixel (6,19), do not take any target label to merge behavior according to algorithm.

6) the 7th scan line, pixel (7,6), this point belong to the target label under its upper left pixel point, check this target label S (3), find equivalent indicia sensing S (4), so this point belongs to target label S (4); In like manner, scanning element S (7,13) also belongs to target label S (4).

7) check target label, S (1), S (2), S (3) significant notation is 0, and only having S (4) significant notation is 1, therefrom can learn, only has a target label in this pattern.

(4) frame selects destination object geometric center location:

Calculate the target label position coordinates, select the marked pixels coordinate of source location by frame, the coordinate figure that will belong to the different picture elements of same connected component averages, and with the position of mean value as the center-of-mass coordinate of this frame favored area or several adjacent, overlapping region, its coordinate Calculation formula is:

$x_c=\underset{i,j\&Element;\mathrm{\&Omega;}}{\mathrm{\&Sigma;}}\frac{j}{N},y_c=\underset{i,j\&Element;\mathrm{\&Omega;}}{\mathrm{\&Sigma;}}\frac{i}{N}$

Wherein: (i, j) is the longitudinal and transverse coordinate of image pixel;

N is total number of picture elements of connected component;

Ω is the set that belongs to the pixel of same connected component;

Implementation is as follows:

The gray value of each picture element of step 1 search bianry image line by line, execution in step 2 is called recursive function when gray value is 0.

Step 2 is set to 1 with the gray scale of picture element, the abscissa of picture element and abscissa and addition, ordinate and ordinate and addition.Begin to search for to 8 communication direction respectively from the present picture element point, judge whether the gray value of these 8 picture elements is 0, if 0 execution in step 2 finishes to return main program otherwise call.

Step 3 is the center-of-mass coordinate of the image of current region position according to the horizontal stroke, mean value ordinate and that calculate the transverse and longitudinal coordinate that return.The next gray value of continuation search is 0 picture element, so obtains the center-of-mass coordinate of all micropores.By the centroid calculation of all pixels of object box favored area, obtain the position coordinates of this regional center point.

Claims (1)

1. based on the little object localization method of pixel, it is characterized in that: may further comprise the steps:

(1) target acquisition

1) videoscanning: fixedly camera scans monitoring to the appointed area in the field range;

2) view image triggers sound and light alarm automatically through the backstage analytical calculation when meeting the default alarm rule;

3) according to alert locations information, carry out alarming determining by the fixing panoramic video of camera;

4) select emphasis scene and target in the video pictures by frame selection tool frame, the frame selection operation triggers the startup of rotary camera;

5) obtain frame selected scenes scape geometric center location point coordinate by hind computation, and according to matching algorithm, calculate and obtain the position of this geometric center in rotary camera, thereby obtain the cloud platform rotation angle;

6) regulate rotation camera head orientation, zoom obtains the HD video of scape position, frame selected scenes;

(2) target localization

1) regulates by the video of rotary camera, obtain the complete HD video picture of localizing objects;

2) carry out the image outline described point at the HD video picture, and by the back-end data processing center, the visual profiles dot information is converted to the sea chart coordinate, project on the sea chart;

3) the projected position point on the sea chart is created as section object by sea chart zone drawing tool, by calculating this regional geometric center, directly obtains the latitude and longitude information of this position at sea chart;

4) by the sea chart referential function, event icons and explanation are added in this position, convenient users at different levels check and share;

5) the target location latitude and longitude information by obtaining arranges the locking monitoring to this target of video camera, and is dynamic in real time to understand it;

6) send event report information to relevant administrative department.

Images