TW202121332A - Method of acquiring detection zone in image and method of determining zone usage - Google Patents

Abstract

A method of acquiring a detection zone in an image comprises by a camera, sequentially acquiring a plurality of images associated with an image capturing scene, by a computing device, computing a plurality of moving traces of a plurality of objects in the images, and displaying the detection zone and another image that is associated with the image capturing scene but different from the images.

Description

Method for obtaining image detection area and judging method for space usage

The present invention relates to a method for obtaining an image detection area and a method for determining space usage, and more particularly to a method for obtaining a detection area and judging space usage based on multiple movement tracks of multiple objects in an image.

In order to monitor the direction of travel, travel time and stay time of vehicles on the road, a variety of monitoring methods have been developed to monitor the behavior of vehicles. However, most monitoring methods are still by setting surveillance cameras to shoot multiple scenes and then monitoring them. The personnel judge whether there are abnormal conditions in these scenes based on the images taken by the surveillance cameras. In addition, whenever a surveillance camera is installed in a new field, it is still necessary to manually judge lanes in different directions, and draw the Region of Interest (ROI) or detection area in each lane accordingly.

However, the artificially drawn ROI may not be the most effective detection area. Therefore, if you calculate or count the traffic flow in this detection area, the statistical data may be inaccurate. In addition, if multiple camera devices need to be installed in a new field, the detection area of the scene shot by each camera device is still manually divided, which causes a waste of time. In addition, the camera device installed in the monitoring field may be affected by external forces (for example, vibration or wind and other force majeure factors), which may cause deviations in the originally set detection area, and the monitoring personnel need to go to A camera device that monitors the field to adjust the offset.

In view of the above, the present invention provides a method for obtaining an image detection area and a method for determining space usage to meet the above requirements, so that when multiple camera devices are installed in a new field, there is no need for monitoring personnel to divide the detection area. This saves time for manual division. In addition, according to the method for determining the use of space according to one or more embodiments of the present invention, after obtaining the detection area, the computing device can determine the flow rate of the object in the capturing scene and whether the object has abnormal behavior, and determine whether the camera has abnormal behavior. Whether the device is offset, the monitoring unit is notified accordingly, and when the camera device is offset, even if the monitoring personnel does not immediately adjust the offset camera device, it will not greatly affect the accuracy of the detection area.

According to an image detection area acquisition method of an embodiment of the present invention, the method includes: sequentially acquiring a plurality of images associated with a capturing scene by a camera; A plurality of movement tracks; a clustering procedure is performed on the movement tracks by the computing device to obtain a detection area; and the detection area is displayed on a display and is related to the imaging scene and is different from the images An image.

According to an embodiment of the space usage determination method of the present invention, the method includes: sequentially obtaining a plurality of images associated with a capturing scene with a camera device; sequentially obtaining images associated with a capturing scene with a camera device A plurality of images; a computing device calculates a plurality of movement trajectories of objects in the images; the computing device performs a clustering procedure on the movement trajectories to obtain a detection area; the computing device is based on the The detection area executes an event detection process, wherein the event detection process uses the computing device to determine whether the behavior of an object to be detected meets an event rule; and the computing device outputs a detection of the event detection process .

The above description of the disclosure and the following description of the implementation manners are used to demonstrate and explain the spirit and principle of the present invention, and to provide a further explanation of the patent application scope of the present invention.

The detailed features and advantages of the present invention are described in detail in the following embodiments. The content is sufficient to enable anyone familiar with the relevant art to understand the technical content of the present invention and implement it accordingly, and is based on the content disclosed in this specification, the scope of patent application and the drawings. Anyone who is familiar with relevant skills can easily understand the purpose and advantages of the present invention. The following examples further illustrate the viewpoints of the present invention in detail, but do not limit the scope of the present invention by any viewpoint.

The method for obtaining the image detection area disclosed in the present invention is used to obtain the detection area of the image capturing scene captured by the camera device. For example, the detection area may be a detection area of an imaging scene on a flat road, a highway, a department store, a shopping mall, a farm, etc., where the imaging scene is preferably a field with multiple moving objects. However, the present invention is not limited to this. For ease of description, the detection area acquisition methods disclosed in the following embodiments will take a road as an example of an image capturing scene.

Please refer to FIGS. 1 and 2 together, in which FIG. 1 is a schematic diagram of a method for obtaining an image detection area according to an embodiment of the present invention; FIG. 2 is an image detection area according to an embodiment of the present invention The flow chart of the acquisition method.

Please refer to step S01: Sequentially obtain multiple images associated with the capturing scene by a camera device. The camera device is, for example, a surveillance video recorder installed on the side of the road. The captured scene is the scene of the road captured by the camera device. , And the image is the image of the road as shown in Figure 1. In detail, the multiple images acquired by the camera device are the images sequentially acquired by the camera device at different time points, and the images include multiple objects O as shown in FIG. 1.

After a plurality of images are obtained by the camera device, the computing device calculates a plurality of movement trajectories MT of the objects in the images in step S03. In detail, the camera device obtains the first image at the first image capturing time, and the computing device uses the neural network deep learning method to identify these objects in the first image, and the positions of these objects in the first image. A plurality of first positions, and the confidence value of the recognition result is higher than a threshold value; the camera device then obtains the second image at the second imaging time, and the computing device uses the neural network deep learning method to recognize the second image These objects, and multiple second positions of the objects in the second image, respectively, and the confidence value of the recognition result is higher than the aforementioned threshold value. Subsequently, the computing device obtains the movement track MT of each object according to the first position and the second position of each object. In other words, the computing device calculates the movement trajectory MT of each object based on the recognition confidence value of each object in the multiple images and the multiple positions of each object based on the neural network deep learning method. The computing device is, for example, the central processing unit of the monitoring unit or the cloud server with computing function, etc.; and the deep learning method of neural network is, for example, the convolutional neural network (Convolutional Neural Network) of artificial intelligence (AI) technology. CNN), however, the present invention is not limited to this.

Please continue to refer to step S03. In detail, when the computing device uses the neural network deep learning method to identify the object in the image, the aforementioned confidence value can be added at the same time, where the confidence value represents the computing device judged by the neural network deep learning method After the object, the degree of certainty of the result of this judgment. When the confidence value reaches the aforementioned threshold value (for example, 70%), it means that the computing device uses the neural network deep learning method to determine the object to be reliable. For example, the computing device uses the neural network deep learning method to identify the object O in the image shown in Figure 1 as a car and generates a confidence value corresponding to this identification. When the confidence value reaches 70%, it means that the computing device uses the neural network. The network deep learning method recognizes that the object O is a car, and the result is reliable.

Please continue to refer to step S03. After the computing device recognizes the object O, the computing device then obtains the coordinate position of the object O in the image according to time sequence. Taking the example of a single object O, the computing device obtains the coordinate position of the object O in each of the multiple images, and connects the coordinate positions in time sequence to obtain the movement track MT of the object O.

After the arithmetic device obtains the movement trajectories MT, the arithmetic device then performs a clustering procedure on the movement trajectories MT to obtain the detection area DZ in step S05, wherein the clustering procedure includes obtaining the data based on the movement trajectories MT by the arithmetic device. There are two boundary points, and the detection area DZ is obtained based on these boundary points. The calculation device obtains the detection zone DZ according to the movement trajectories MT based on a probability distribution function. For the convenience of subsequent description, the normal distribution function is taken as an example below. The normal distribution function is preferably a binary normal distribution function. And one dimension of the binary normal distribution function is the density of the movement trajectory MT, and the other dimension is the distribution of the intersection of the movement trajectory MT and a reference line.

Please refer to FIG. 3(a). FIG. 3(a) is a schematic diagram of a method for obtaining an image detection area according to an embodiment of the present invention. In detail, the implementation manner of performing a clustering procedure on the movement trajectories MT to obtain boundary points may be that the computing device summarizes the normal distribution function based on the movement trajectories MT, and uses the two boundary values of the confidence interval of the normal distribution function respectively. Define the boundary line of the detection zone DZ, where the confidence interval is, for example, 68% to 95% of the normal distribution function, but the present invention is not limited to this. In particular, according to the implementation of the normal distribution function, the normal distribution functions of the first intersection N1 and the two first boundary points BP1 of the confidence interval are obtained, and the normal distribution functions of the second intersection N2 and The two second boundary points BP2 of the confidence interval, and the area enclosed by the two first boundary points BP1 and the second boundary point BP2 is used as the detection zone DZ. Wherein, the aforementioned first intersections N1 are the intersections of the movement trajectories MT and the reference line L1, and the aforementioned second intersections N2 are the intersections of the movement trajectories MT and the reference line L2. In particular, each of the movement trajectories MT extends from its starting point to its end, and the first intersection N1 preferably includes the starting point of one of the movement trajectories MT, and the second intersection N2 preferably includes the The end point of one of these moving tracks MT.

Please refer to Figure 3(b). Figure 3(b) is a schematic diagram of another method for obtaining the boundary line BL of the detection zone DZ. The method for obtaining the boundary line BL is based on at least two groups of movement trajectories MT, namely The normal distribution function formed by the intersection of these movement trajectories MT and a single reference line has at least two confidence intervals. For ease of understanding, only the movement trajectory MT that can form two confidence intervals is described below. After obtaining the two confidence intervals corresponding to the reference line L1 and the two confidence intervals of the reference line L2, the computing device is based on the two confidence intervals of the reference line L1 and the reference line L2 based on the movement trajectories MT of the same group among the movement trajectories MT. The two midpoints of the interval obtain the midline CL1, and the midline CL2 is obtained based on the two midpoints of the two confidence intervals of the reference line L1 and the reference line L2 based on the movement trajectory MT of the other group among the movement trajectories MT. Obtain the boundary line BL of the detection area based on the second center line CL1 and CL2. The arithmetic device respectively summarizes the normal distribution function based on the movement trajectories MT of the two groups. The realization method can be obtained by the reference line L1 and L2 as mentioned above. Therefore, the operation details of the normal distribution function are not summarized here. Go into details.

In detail, please refer to the group of movement trajectories MT on the left in Figure 3(b). After the computing device summarizes the two normal distribution functions based on the movement trajectories MT, the computing device uses the confidence interval of the two normal distribution functions The connection between the first vertex CP1 and the second vertex CP2 of is used as the first centerline CL1, where the first vertex CP1 and the second vertex CP2 are respectively 50% of the two normal distribution functions, but the present invention is not limited to this. Please refer to another group of movement trajectories MT on the right in Figure 3(b). Similarly, the computing device uses the connection between the first vertex CP1' and the second vertex CP2' of the confidence interval of the other two normal distribution functions as the second Centerline CL2. Next, the arithmetic device equally divides the distance between the first center line CL1 and the second center line CL2 to obtain the boundary line BL, so that the distance between any point on the boundary line BL and the first center line CL1 is equal to that on the boundary line BL The distance between this arbitrary point and the second center line CL2. In addition, after obtaining the first vertex CP1/CP1' and the second vertex CP2/CP2', the computing device can also obtain the first intermediate point BP1 between the first vertex CP1 and the second vertex CP2, and the first vertex CP1' The second intermediate point BP2 between the second vertex CP2' and the second intermediate point BP2, and a line segment connecting the first intermediate point BP1 and the second intermediate point BP2 is used as the boundary line BL.

Please continue to refer to the embodiment of FIG. 3(b) above. After the boundary line BL is obtained, the computing device can symmetrical the boundary line BL to another of the first center line CL1 by taking the first center line CL1 as the center line. One side is used to obtain another boundary line, and the detection area DZ is enclosed by two boundary lines BL and two reference lines L1 and L2. Similarly, the computing device can also use the second center line CL2 as the center line to obtain the detection zone DZ in a symmetrical manner. It should be noted that when the first center line CL1 and the second center line CL2 are respectively used as center lines to obtain three boundary lines BL, the computing device can direct the two boundary lines BL located on both sides of the first center line CL1 toward The first center line CL1 is moved by a predetermined distance as the updated second boundary line BL; and the second boundary lines BL located on both sides of the second center line CL2 are moved toward the second center line CL2 by the predetermined distance to be updated The second boundary line BL, where the preset distance can be 20% of the distance between the two first vertices CP1 and CP1', or 20% of the distance between the two second vertices CP2 and CP2', the The invention is not limited to this. According to this, the two detection areas DZ that are originally connected can be adjusted to be separated from each other.

In addition, if there are at least four groups of movement trajectories MT in the image, the computing device can execute the method of obtaining the boundary line BL with a normal distribution function as disclosed in FIG. 3(b), and obtain two boundaries based on the four groups of movement trajectories MT. Line BL, and then circle the detection area DZ with two boundary lines BL and two reference lines L1 and L2.

According to the method of obtaining the boundary line shown in FIGS. 3(a) and 3(b), the computing device can obtain the detection zone DZ based on the boundary line. In the example of FIG. 1, the detection area DZ is in the lane where the vehicle is traveling, and the boundary line BL is overlapped or adjacent to the lane line dividing the lanes.

After the detection area DZ is obtained by the computing device, the method for obtaining the detection area DZ of this embodiment then displays the detection area DZ on the display in step S07 and another image associated with the image capturing scene and different from the images, wherein The other image is preferably a real-time image applied to the detection area. The display is, for example, a display screen provided in a monitoring unit, but the present invention is not limited to this.

In detail, after obtaining the detection zone DZ, the camera device continuously obtains real-time images of the capturing scene for display on the display, which means that the time point when the camera device obtains these images is earlier than the time point when the camera device obtains the real-time image, and The real-time image displayed on the display is marked with a detection zone DZ, wherein the detection zone DZ marked on the real-time image may be one or more, and the present invention is not limited to this.

After the display shows the detection zone DZ and the real-time image (step S07), the implementation of this embodiment may further include obtaining a plurality of images sequentially by the camera device again (step S01) every time an interval has elapsed (step S01), in order to compare different The traffic condition of the time period obtains the corresponding detection zone DZ.

Please refer to FIG. 4, which is a flowchart of a method for obtaining a detection area according to another embodiment of the present invention. The steps S01 to S05 and steps S01' to S05' disclosed in FIG. 4 are the same as the steps S01 to S05 disclosed in FIG. 2, so the same operation details will not be repeated here. However, in the detection area obtaining method disclosed in FIG. 4, the multiple images obtained in step S01 are different from the multiple images obtained in step S01', and the movement trajectory calculated in step S03 is also different from that calculated in step S03' Therefore, the first detection area obtained after performing the clustering procedure on the movement trajectories calculated in step S03 (step S05) is different from performing the clustering on the movement trajectories calculated in step S03'. The second detection area obtained after the similar procedure (step S05'). In addition, in this embodiment, it is preferable that steps S01 to S05 are executed first, and steps S01' to S05' are executed later.

In detail, after obtaining the first detection area in step S05 and displaying the first image with the first detection area in step S07, the computing device then executes steps S01' to S05' to obtain the second detection area, where The first detection area and the second detection area may be one or more, and the invention is not limited thereto. In other words, the first image and the second image described in steps S01 and S01' are preferably real-time images of the same capturing scene but at different time points, respectively.

After obtaining the first detection area and then obtaining the second detection area, the computing device compares the first detection area with the second detection area in step S09 to obtain a comparison value, wherein the comparison value obtained by the computing device in step S09 The value is used to indicate the amount of overlap between the first detection area and the second detection area. In detail, the first detection area and the second detection area that are related to each other are preferably the detection areas obtained on different dates but the same time period, and the computing device compares the first detection areas on different dates but the same time period. Area and the second detection area to determine whether the camera device is offset, and to obtain the overlap between the first detection area and the second detection area can be achieved by judging that the second detection area overlaps the first detection area. The ratio value of the detection area, and use this ratio value as the comparison value.

Please continue to refer to step S11, the arithmetic device then judges whether the comparison value is lower than an overlap threshold in step S11. In other words, the arithmetic device determines whether the camera device itself has an offset based on the comparison value and the overlap threshold value, where the overlap threshold value is, for example, 80%, but the present invention is not limited to this.

When the computing device determines in step S11 that the computing device determines that the comparison value is lower than the overlap threshold, it means that the camera itself may be offset due to external forces such as wind, vibration, etc., and the computing device then uses the second detection area in step S13 The detection area is updated, and a notification is output in step S15 for display on the display. On the contrary, if the comparison value is not lower than the overlap threshold value, it means that the camera device has no offset, or the offset of the camera device is within the allowable range, and the computing device returns to step S09 and continues to perform the first detection area and the second detection area. Second, the detection area is compared.

It should be noted that step S13 shown in FIG. 4 is performed before step S15, but step S13 and step S15 can also be performed at the same time, or step S13 is performed after step S15, and the present invention is not limited thereto.

In other words, the computing device updates the second detection area in step S13. The detection area uses the newly acquired second detection area as the detection area, and outputs a notification in step S15 to notify the surveillance unit that the camera device may have abnormal conditions . Accordingly, when the camera device is shifted, even if the worker does not adjust the camera device in real time, the detection area of the image capturing scene will not be affected.

Please refer to FIG. 5, which is a flowchart of a method for determining space usage according to an embodiment of the present invention. The method for determining the use of space as disclosed in FIG. 5 is preferably used to determine whether the behavior of the object to be detected conforms to an event rule. The steps S01 to S07 disclosed in FIG. 5 are the same as the steps S01 to S07 disclosed in FIG. 2, so the steps S01 to S07 disclosed in FIG. 5 will not be repeated here. However, the violation determination method based on the detection area disclosed in FIG. 5 is that after the computing device executes the clustering procedure to obtain the detection area (step S05), the computing device can display the detection area on the display in step S07 and be associated with the image capture. In addition to another image that is different from the scene, the computing device can also execute an event detection process based on the detection area in step S08 to determine whether the behavior of an object to be detected conforms to an event rule. However, the above steps S07 and S08 can also be performed together. Among them, the event rules are, for example, the traffic of objects in the detection area, no objects in the detection area (forbidden area detection), and whether any object in the detection area stays for a predetermined period of time ( Parking detection), whether any object in the detection area moves in a preset direction (reverse detection), and whether the moving speed of any object in the detection area falls within a preset speed zone (overspeed /Slow detection), but the present invention is not limited to this.

In detail, if the event rule is that no object can exist in the detection area, executing the event detection procedure includes determining whether the coordinate position of the object to be detected falls within the detection area. Similarly, if the event rule of the event detection process is about the stay time or movement of any object in the detection area, and multiple coordinate positions of the object to be detected fall within the detection area, then the event detection will be executed. In addition to the multiple coordinate positions of the object to be detected, the measurement procedure also includes multiple time information corresponding to the coordinate positions to determine the dwell time of the object to be detected based on multiple coordinate positions and multiple time information Or move.

Please continue to refer to step S08. After the computing device executes the event detection process in step S08, the computing device outputs the detection result of the event detection process in step S10. In detail, if the event detection process is to obtain the object flow in the detection area (for example, the traffic flow), the detection result preferably includes the object flow in the detection area or an abnormal condition of the object flow. In addition, the computing device may be a memory device that outputs the detection result to the monitoring unit for recording, or output to the monitor of the monitoring unit for display, etc. The present invention is not limited to this.

The detection result may also include when it is determined that the behavior of the object to be detected does not conform to the event rule, the computing device outputs a notification for display on the display. Conversely, when the computing device determines in step S08 that the behavior of the object to be detected conforms to the event rule, it returns to step S01 to sequentially obtain multiple images associated with the capturing scene by the camera to update the detection area.

Taking the example of the above vehicle, the object to be detected is the vehicle that exists in the imaging scene after the detection area is obtained. Therefore, if the coordinate position of the vehicle (the object to be detected) falls within the detection area, the vehicle is in the detection area. The stay time in the detection area exceeds the preset time of the event rule, the moving direction of the vehicle in the detection area does not meet the preset direction of the event rule (for example, reverse driving), or the moving speed of the vehicle in the detection area falls below Outside the preset speed section, the computing device will output a notification to the display to remind the monitoring unit that there may be an abnormal situation in the capturing scene.

Please continue to refer to step S08. However, when the computing device executes the event detection procedure in step S08 to determine that the behavior of the object to be detected does not conform to the event rule, it can also return to step S01 to sequentially obtain the association with the capturing scene by the camera When the behavior of the object to be detected meets the event rule, the computing device outputs a notification for display on the display, and the present invention is not limited to this.

In summary, according to the detection area acquisition method of one or more embodiments of the present invention, the effective detection area can be divided based on the movement trajectory of the object, and it is not necessary to install multiple camera devices in a new field. The detection area needs to be divided by the monitoring personnel to save the time of manual division. In addition, according to the method for determining the use of space according to one or more embodiments of the present invention, after obtaining the detection area, the computing device can determine the flow rate of the object in the capturing scene and whether the object has abnormal behavior, and determine whether the camera has abnormal behavior. Whether the device is offset, the monitoring unit is notified accordingly, and when the camera device is offset, even if the monitoring personnel does not immediately adjust the offset camera device, it will not greatly affect the accuracy of the detection area.

Although the present invention is disclosed in the foregoing embodiments, it is not intended to limit the present invention. All changes and modifications made without departing from the spirit and scope of the present invention fall within the scope of the patent protection of the present invention. For the scope of protection defined by the present invention, please refer to the attached scope of patent application.

O: Object MT: Movement track DZ: detection zone L1, L2: reference line N1: the first node N2: second intersection BP1: the first boundary point BP2: Second boundary point CP1, CP1’: the first vertex CP2, CP2’: the second vertex CL1: First midline CL2: Second midline BL: boundary line

FIG. 1 is a schematic diagram of a method for obtaining an image detection area according to an embodiment of the present invention. FIG. 2 is a flowchart of a method for obtaining an image detection area according to an embodiment of the invention. 3(a) and 3(b) are schematic diagrams of a method for obtaining an image detection area according to an embodiment of the present invention. FIG. 4 is a flowchart of a method for obtaining an image detection area according to another embodiment of the present invention. FIG. 5 is a flowchart of a violation determination method according to an embodiment of the present invention.

Claims (10)

A method for obtaining a detection area, the method comprising: sequentially obtaining a plurality of images associated with a capturing scene by a camera device; calculating a plurality of movement tracks of a plurality of objects in the images by a computing device; The computing device executes a clustering process on the movement tracks to obtain a detection area; and displays the detection area and another image related to the capturing scene and different from the images on a display. The method for obtaining the detection area according to claim 1, wherein the clustering procedure comprises: obtaining two boundary points of a confidence interval of a probability distribution function based on the intersection of the movement trajectories and a reference line by the computing device, based on the The intersection of these movement trajectories and another reference line obtains the two boundary points of the confidence interval of another probability distribution function, and the area enclosed by the four boundary points is used as the detection area. The method for obtaining a detection area according to claim 1, wherein the method comprises: obtaining a first detection area and a second detection area that are related to each other by comparing the clustering procedure twice with the computing device Obtain a comparison value, wherein the acquisition time for acquiring the multiple images of the first detection area is earlier than the acquisition time for acquiring the other multiple images of the second detection area; when the computing device determines the comparison When the value is lower than an overlap threshold, update the detection area with the second detection area; and output a notification for the display to display. The method for obtaining the detection area according to claim 1, wherein calculating the movement trajectories of the objects in the images by the computing device includes: using the computing device according to a first image capturing time of the objects A plurality of first positions and a plurality of second positions of the objects at a second imaging time calculate the movement trajectories. The detection area obtaining method according to claim 1, wherein the method further comprises: recognizing the objects in the images by a neural network (Neural Network) deep learning method and obtaining a confidence associated with the objects Use the computing device to determine whether the confidence value reaches a threshold; and when the confidence value reaches the threshold, use the computing device to calculate the movement trajectories of the objects in the images. A method for judging space usage, the method comprising: sequentially obtaining a plurality of images associated with a capturing scene with a camera device; calculating a plurality of movement trajectories of a plurality of objects in the images with an arithmetic device; The computing device executes a clustering process on the movement tracks to obtain a detection area; the computing device executes an event detection process based on the detection area, wherein the event detection process uses the computing device to determine a waiting area Detect whether the behavior of the object conforms to an event rule; and output a detection result of the event detection procedure by the computing device. The method for determining the use of space according to claim 6, wherein the clustering procedure includes: obtaining the two boundary points of the confidence interval of a probability distribution function by the computing device based on the intersection of the movement trajectories and a reference line, based on The intersection of the movement trajectories and another reference line obtains the two boundary points of the confidence interval of another probability distribution function, and the area enclosed by the four boundary points is used as the detection area. The method for determining space usage according to claim 6, wherein the event rule is whether any object in the detection area stays for a predetermined period of time, and the computing device determines the to-be-detected based on the detection area Whether the behavior of the object conforms to the event rule includes: judging by the computing device based on the detection area whether the position of a mark of the object to be detected falls within the detection area; and when judging the coordinate position of the object to be detected When it falls within the detection area, the computing device determines whether the time that the coordinate position of the object to be detected falls within the detection area reaches the preset time. The method for determining space usage according to claim 6, wherein the event rule is whether any object in the detection area moves in a predetermined direction, and the computing device determines the to-be-detected based on the detection area Whether the behavior of the object conforms to the event rule includes: judging by the computing device based on the detection area whether the position of a mark of the object to be detected falls within the detection area; and when judging the coordinate position of the object to be detected When falling within the detection area, the computing device determines whether the object to be detected moves along the predetermined direction based on the multiple coordinate positions of the object to be detected and the corresponding multiple time information. The method for determining space usage according to claim 6, wherein the event rule is whether the moving speed of any object in the detection area falls within a predetermined speed range, and the computing device is based on the detection The area determining whether the behavior of the object to be detected conforms to the event rule includes: judging by the computing device based on the detection area whether the position of a mark of the object to be detected falls within the detection area; and when determining the object to be detected When the coordinate position of the object to be detected falls within the detection area, the computing device is used to determine whether the moving speed of the object to be detected falls within the detection area based on the multiple coordinate positions of the object to be detected and the corresponding multiple time information. Within the preset speed section.

Images