Adaptive monitoring identification method, system and device for specified area and medium

Abstract

The invention provides a self-adaptive monitoring identification method, system and device for a specified area and a medium. The method comprises the following steps: acquiring a panoramic picture of a monitored area through a camera; marking a region of interest in the panoramic picture, and obtaining boundary point coordinates of the region of interest; extracting feature points from the panoramic picture by using a feature point extraction algorithm; drawing the feature points on the panoramic picture, screening out static feature points, and generating a first feature point set based on the static feature points; acquiring a real-time monitoring picture through a camera, extracting feature points, and generating a second feature point set; performing feature matching on the second feature point set and the first feature point set, calculating a homography matrix, and establishing a mapping relation between the real-time monitoring picture and the panoramic picture; based on the mapping relationship, mapping the boundary point coordinates of the region of interest into the real-time monitoring picture, and generating the mapping coordinates of the boundary points; and judging whether the position of the camera needs to be adjusted according to the mapping coordinates of the boundary points and the range of the real-time picture.

Description

Technical Field

[0001] The present invention relates to the field of image processing technology, and more particularly to an adaptive monitoring and recognition method, system, device and medium for a designated area. Background Art

[0002] With the widespread popularity of cameras, many places have installed cameras to monitor specific areas of interest. However, these areas of interest are often not the entire range captured by the camera, but only a part of it. In specific scenarios such as monitoring multiple workstations separately and monitoring entrances and exits, accurate distinction of areas is particularly important. At the same time, cameras usually have horizontal and vertical rotation functions and zoom operations, which may cause the position and size of the area of ​​interest in the overall shooting area to change. When these changed areas of interest need to be further processed, many challenges will be encountered.

[0003] At present, the main method to solve this problem is to first determine the initial state of a camera, such as a specific horizontal deflection angle, vertical deflection angle, and zoom position. Once the camera is changed, it is usually necessary to manually adjust it back to the initial state. However, most monitoring systems still rely on manual configuration of camera parameters, making it difficult to achieve automatic recognition and dynamic tracking of the area of ​​interest. This not only increases the complexity of operation, but also severely limits the response speed and flexibility of the monitoring system.

[0004] In addition, when dealing with complex and ever-changing monitoring scenarios, existing systems often have problems with blind spots or waste of resources, making it difficult to achieve efficient and accurate monitoring results. Existing technologies are mainly based on monitoring systems with fixed cameras, which monitor through preset monitoring ranges and fixed camera parameters. Although they can meet basic monitoring needs, these systems are unable to meet actual needs when faced with scenes where the location and size of the area of ​​interest change frequently or where high-precision monitoring is required. Summary of the invention

[0005] In view of the above problems, the purpose of the present invention is to provide an adaptive monitoring and identification method, system, device and medium for a specified area. Through the adaptive monitoring and identification method, comprehensive, accurate and intelligent monitoring of the monitored area is achieved, the efficiency and accuracy of the monitoring system are improved, and the flexibility and adaptability of the monitoring system are enhanced.

[0006] In order to achieve the above object, the present invention is implemented through the following technical solutions: In a first aspect, the present invention discloses an adaptive monitoring and identification method for a specified area, comprising: By adjusting the camera to obtain the video stream of the monitoring area, and then obtaining the monitoring picture, and by splicing the monitoring pictures, obtaining the panoramic picture of the monitoring area; Use the labelme tool to mark the area of ​​interest in the panoramic image and obtain the coordinates of the boundary points of the area of ​​interest; Use feature point extraction algorithm to extract feature points from the panoramic image; Draw the feature points on the panoramic image, and use the labelme tool to draw the convex area where the static object is located. There can be multiple convex areas. Use the cross multiplication method to determine that the feature points in the marked convex area are static feature points, and generate the first feature point set based on the static feature points. Acquire a real-time monitoring picture through a camera, and extract feature points from the real-time monitoring picture using a feature point extraction algorithm to generate a second feature point set; Performing feature matching on the second feature point set and the first feature point set, calculating a homography matrix, and establishing a mapping relationship between the real-time monitoring image and the panoramic image according to the homography matrix; Based on the mapping relationship, the coordinates of the boundary points of the focus area are mapped to the real-time monitoring image to generate the mapping coordinates of the boundary points; According to the mapping coordinates of the boundary points and the range of the real-time monitoring picture, it is determined whether the position of the camera needs to be adjusted. If so, the movement angle of the camera is determined according to the mapping coordinates of the boundary points, and the camera movement is controlled to perform video acquisition.

[0007] Furthermore, the method of acquiring a video stream of the monitoring area by adjusting the camera, and then acquiring a monitoring picture, and acquiring a panoramic picture of the monitoring area by stitching the monitoring pictures, includes: Move the camera to its extreme position so that the camera's viewing angle can maximize the monitoring area and collect video streams; For the video stream collected at each extreme position, monitoring pictures are obtained by frame extraction, and the image stitching algorithm is applied to combine the monitoring pictures into a panoramic picture.

[0008] Furthermore, the feature points are drawn on the panoramic image, and the convex area where the static object is located is drawn by the labelme tool. There may be multiple convex areas. The feature points in the marked convex area are judged as static feature points by using the cross multiplication discrimination method. The first feature point set is generated based on the static feature points, including: Use the Pillow software tool to draw the feature points on the panoramic image, use the labelme tool to mark the feature points on the static object with a convex polygon frame, and obtain the vertex coordinates of the convex polygon to generate a convex polygon vertex set ; Based on the convex polygon vertex set, the cross product discrimination method is used to determine whether any feature point t is a static feature point. If so, the feature point t is recorded in the first feature point set; the coordinates of the feature point t are . Further, the cross product determination method is used to determine whether any feature point t is a static feature point, including: Calculate the values ​​of the following discriminants respectively:

[0009]

[0010] If the values ​​of the discriminant are all positive or all negative, the feature point t is a static feature point.

[0011] Further, the feature matching of the second feature point set with the first feature point set, calculating the homography matrix, and establishing a mapping relationship between the real-time monitoring picture and the panoramic picture according to the homography matrix includes: Based on Lowe's algorithm, feature matching is performed on the second feature point set and the first feature point set to determine four pairs of matching points; Each pair of matching points There are the following conversion relationships:

[0012]

[0013] in, is the feature point in the first feature point set, and its coordinates are , is the feature point in the second feature point set, and its coordinates are , is the homography matrix; Based on the above transformation relationship and four pairs of matching points, the value of the homography matrix is ​​calculated; The mapping relationship between the real-time monitoring image and the panoramic image is:

[0014] in, The coordinates of the boundary points of the panoramic image focus area. It is the mapping coordinates of the boundary points in the monitoring image.

[0015] Furthermore, mapping the boundary point coordinates of the focus area to the real-time monitoring picture based on the mapping relationship to generate the mapping coordinates of the boundary points includes: The mapping coordinates of the boundary points are calculated according to the following formula: .

[0016] Further, judging whether the position of the camera needs to be adjusted according to the mapping coordinates of the boundary points and the range of the real-time image, and if so, determining the moving angle of the camera according to the mapping coordinates of the boundary points, includes: Set the minimum detection pixel size of the camera to , then the zoom factor P is: ; in, To monitor the actual pixel size; judge , , , Whether one of the four inequalities holds true; If so, you need to adjust the position of the camera; Based on the mapping coordinates of the boundary points, the four extreme points of the circumscribed rectangle of the corresponding area are calculated using the following formula:

[0017] The center coordinates of the circumscribed rectangle are calculated using the following formula: , ; According to the center coordinates of the monitoring image and the center coordinates of the bounding rectangle The relative position of determines the moving direction of the camera; The horizontal and vertical movement angles of the camera are calculated using the following formula:

[0018]

[0019] in, is the horizontal movement angle of the camera, is the vertical movement angle of the camera. In a second aspect, the present invention further discloses an adaptive monitoring and identification system for a designated area, comprising: A panoramic picture acquisition module is used to acquire the video stream of the monitoring area by adjusting the camera, and then acquire the monitoring picture, and acquire the panoramic picture of the monitoring area by splicing the monitoring pictures; The module for extracting coordinates of the region of interest is used to mark the region of interest in the panoramic image using the labelme tool and obtain the coordinates of the boundary points of the region of interest; A feature point extraction module is used to extract feature points from the panoramic image using a feature point extraction algorithm; The static feature point screening module is used to draw feature points on the panoramic image and use the labelme tool to draw the convex area where the static object is located. There can be multiple such convex areas. The cross multiplication method is used to determine that the feature points in the marked convex area are static feature points, and the first feature point set is generated based on the static feature points. A real-time image feature point extraction module is used to obtain a real-time monitoring image through a camera, and extract feature points from the real-time monitoring image using a feature point extraction algorithm to generate a second feature point set; A feature matching and mapping module, used to perform feature matching on the second feature point set and the first feature point set, calculate a homography matrix, and establish a mapping relationship between the real-time monitoring image and the panoramic image according to the homography matrix; A coordinate mapping module is used to map the coordinates of the boundary points of the focus area to the real-time monitoring picture based on the mapping relationship to generate the mapping coordinates of the boundary points; The position recognition and adjustment module is used to determine whether the position of the camera needs to be adjusted based on the mapping coordinates of the boundary points and the range of the real-time monitoring image. If so, the camera movement angle is determined based on the mapping coordinates of the boundary points, and the camera movement is controlled to perform video acquisition.

[0020] In a third aspect, the present invention further discloses an adaptive monitoring and identification device for a designated area, comprising: A memory, used for storing an adaptive monitoring and recognition program for a designated area; A processor is used to implement the steps of the method for adaptive monitoring and identification of a designated area as described in any one of the above items when executing the adaptive monitoring and identification program of the designated area.

[0021] In a fourth aspect, the present invention further discloses a readable storage medium, on which is stored an adaptive monitoring and identification program for a specified area. When the adaptive monitoring and identification program for a specified area is executed by a processor, the steps of the adaptive monitoring and identification method for a specified area as described in any one of the above items are implemented.

[0022] Compared with the prior art, the present invention has the following beneficial effects: 1. The present invention uses an adaptive monitoring and recognition method to adjust the camera position in real time, ensure that the monitoring area always remains within the field of view, and accurately identify changes in the area of ​​interest, thereby improving the efficiency and accuracy of monitoring.

[0023] 2. The present invention obtains a panoramic picture by splicing monitoring pictures, which can fully understand the situation of the monitoring area. At the same time, by using feature point matching and homography matrix to establish a mapping relationship, the boundary point coordinates of the area of ​​interest can be accurately mapped to the real-time monitoring picture to achieve detail capture.

[0024] 3. The present invention uses the cross-product discrimination method to determine the static properties of feature points, and intelligently determines whether the camera position needs to be adjusted based on the comparison between the mapped coordinates of the boundary points and the real-time image range. Once adjustment is required, the camera's moving angle can be automatically calculated to achieve dynamic adjustment, ensuring the continuity and accuracy of monitoring.

[0025] 4. The camera adjustment strategy of the present invention is based on real-time data and algorithm calculation, so that the monitoring system can flexibly respond to various monitoring scenarios and demand changes, thereby improving the adaptability and flexibility of the monitoring system.

[0026] 5. The present invention reduces the steps of human intervention and manual operation, reduces the difficulty and complexity of operation, and improves work efficiency through automated monitoring, identification and adjustment processes.

[0027] It can be seen that compared with the prior art, the present invention has outstanding substantive features and significant progress, and the beneficial effects of its implementation are also obvious. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0029] Figure 1 It is a method flow chart of a specific implementation mode of the present invention.

[0030] Figure 2 It is a system structure diagram of a specific implementation mode of the present invention. DETAILED DESCRIPTION

[0031] In order to enable those skilled in the art to better understand the scheme of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific implementation methods. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0032] See also Figure 1 As shown, this embodiment provides an adaptive monitoring and identification method for a specified area, including the following steps: S1: The video stream of the monitoring area is obtained by adjusting the camera, and then the monitoring picture is obtained. The panoramic picture of the monitoring area is obtained by stitching the monitoring pictures.

[0033] In a specific implementation, the video stream is acquired by adjusting the camera for the all-round area that can be monitored, and the monitoring picture is acquired by frame extraction; the acquired pictures are spliced ​​to obtain a panoramic picture that the camera can take, which can be achieved by deformity correction, image matching, spherical projection, computational stitching, and image fusion.

[0034] As an example, the specific implementation process of this step is as follows: S101. Camera positioning and adjustment: In order to build a solution that can fully capture and stitch together a panoramic image, first, ensure that the camera can move freely to the four extreme positions within its physical reach: upper left corner, upper right corner, lower left corner, and lower right corner. This usually requires the camera to have a controllable device such as an electric pan / tilt or a robotic arm to achieve precise position adjustment. At each extreme position, fine-tune the camera's pitch angle, rotation angle, and zoom level through software or manual methods to ensure that the target area can be covered to the maximum extent at each viewing angle, while reducing overlaps and blind spots.

[0035] S102, Video stream acquisition: Start the camera and configure it to output real-time video streams with appropriate resolution and frame rate. Select an appropriate video encoding format (such as H.264 or H.265) to ensure the stability and transmission efficiency of the video stream. Use professional video stream processing libraries (such as OpenCV, FFmpeg, etc.) to receive and process these video stream data to ensure the continuity and integrity of the data.

[0036] S103, image selection and sorting: From the frames extracted at each position, select a group of the most representative images based on image quality (such as clarity, exposure), consistency of shooting angles, time proximity, no obstruction, etc. Sort these images according to the logical order of camera movement (upper left → upper right → lower right → lower left or adjusted according to the actual layout) to prepare for subsequent image stitching.

[0037] S104, panorama stitching: Apply image stitching algorithms (such as distortion correction, image matching, spherical projection, computational stitching, image fusion, etc.) to combine the selected images into a complete panoramic image. This step can use common image processing algorithms to ensure that the edges of the stitched panoramic image are smoothly transitioned without obvious dislocation or ghosting. Finally, make final adjustments to the stitched panoramic image, including color balance, brightness adjustment, and possible cropping, to achieve the best visual effect.

[0038] S2: Use the labelme tool to mark the area of ​​interest in the panoramic image and obtain the coordinates of the boundary points of the area of ​​interest.

[0039] In a specific implementation manner, the panoramic image is circled with areas of interest such as workstations, doorways, and driving seats, and the labelme tool is used to mark them, and the coordinates of the boundary points of the areas of interest are obtained.

[0040] S3: Extract feature points from the panoramic image using a feature point extraction algorithm.

[0041] In this step, a feature point extraction algorithm (such as SIFT, SURF, ORB, etc.) is used to extract significant and stable feature points from the panoramic image.

[0042] S4: Draw the feature points on the panoramic image, and use the labelme tool to draw the convex area where the static object is located. There can be multiple such convex areas. Use the cross multiplication method to determine that the feature points in the marked convex area are static feature points, and generate the first feature point set based on the static feature points.

[0043] In a specific implementation, since some objects are movable, feature points on or at the edge of such objects are discarded. Therefore, some areas that do not contain movable objects can be determined in the panoramic image. If there are feature points in the area, the feature points are retained in the feature point set, and the feature point set at this location is recorded as the first feature point set. The boundary coordinates and feature point set of the focus area are retained for later use.

[0044] As an example, first use the Pillow software tool to draw the feature points on the panoramic image, use the labelme tool to frame the feature points on the static object with a convex polygon, and obtain the vertex coordinates of the convex polygon to generate a convex polygon vertex set. Then, based on the convex polygon vertex set, the cross product discrimination method is used to determine whether any feature point t is a static feature point. If so, the feature point t is recorded in the first feature point set; the coordinates of the feature point t are . Among them, the specific formula of the cross-product discriminant method is as follows:

[0045]

[0046] It should be noted that if the results of the above n equations have the same sign, it means that the feature point t is inside the convex polygon, that is, the feature point is a static feature point. Similarly, by calculating all the feature points through the above steps, all the static feature points can be found.

[0047] Finally, based on the obtained static feature points, a first feature point set is generated.

[0048] S5: Acquire a real-time monitoring picture through a camera, and use a feature point extraction algorithm to extract feature points from the real-time monitoring picture to generate a second feature point set.

[0049] In a specific implementation, a real-time monitoring picture is obtained by collecting a video stream through a camera. Among them, OpenCV and FFmpeg can be used to extract frames to obtain a real-time monitoring picture. Then, the real-time monitoring picture is subjected to SIFT feature monitoring to obtain feature points of the picture and generate a second feature point set.

[0050] It should be noted that in this step, there is no need to select a static feature point set.

[0051] S6: Perform feature matching on the second feature point set and the first feature point set, calculate a homography matrix, and establish a mapping relationship between the real-time monitoring image and the panoramic image according to the homography matrix.

[0052] In a specific implementation, a feature point match between the first feature point set and the second feature point set is established, a homography matrix is ​​obtained, and a mapping relationship between the first feature point set and the second feature point set is established based on the homography matrix, that is, a mapping relationship between the corresponding images. Finally, the initial boundary coordinates of the region of interest are mapped to the new image based on the established mapping relationship.

[0053] As an example, the static feature points and the newly obtained feature point set can be matched using the matching based on Lowe's algorithm. The specific formula is: Define the homography matrix, simplified notation:

[0054] in, is the feature point in the first feature point set, is the feature point in the second feature point set, It is the homography matrix (3*3 matrix), which contains the camera intrinsic parameter matrix, rotation translation and plane parameter information. are two corresponding points in different images. After introducing the homography matrix, the points of the a series pixels can be directly mapped to the points of the b series pixels. The specific form is solved below The element is expanded.

[0055] The homography matrix is ​​calculated by matching feature points. The definition of the homography matrix is ​​calculated by rotation and translation information. In reality, sometimes the rotation and translation information is unknown, but the matching points in the two images are known, and the homography matrix can be calculated from the matching points. For a pair of matching points on the image, there is the following relationship:

[0056]

[0057] Since it involves matrix transformation, it can be normalized as follows.

[0058]

[0059] Now there are only 8 unknowns, and a pair of matching points provides 2 equations, which are as follows:

[0060] Now we need 4 pairs of points to find 8 variables. The 8 equations corresponding to the 4 pairs of points are:

[0061] The above equation can solve all the variables through the simplest Gaussian elimination method, that is, the homography matrix is ​​obtained, and the corresponding mapping relationship is also obtained.

[0062] Based on the formula , the formula for the mapping relationship can be determined as:

[0063] in, The coordinates of the boundary points of the panoramic image focus area. It is the mapping coordinates of the boundary points in the monitoring image.

[0064] S7: Mapping the coordinates of the boundary points of the focus area to the real-time monitoring image based on the mapping relationship to generate mapping coordinates of the boundary points.

[0065] In a specific implementation, the coordinates of the border points of the panoramic image focus area are , then the mapped coordinates are .

[0066] Substituting into formula (1), we can get the mapping coordinates of the boundary points:

[0067] in, They are all known numbers.

[0068] S8: judging whether the position of the camera needs to be adjusted according to the mapping coordinates of the boundary points and the range of the real-time monitoring image, and if so, determining the moving angle of the camera according to the mapping coordinates of the boundary points, controlling the movement of the camera and performing video acquisition.

[0069] In a specific implementation, if the monitoring area of ​​interest is not in the surveillance image, a mapping relationship can be established through the mapping relationship established above and the movement of the camera, and the distance that the camera needs to rotate horizontally and vertically is converted into the angle and direction that needs to be adjusted, and the angle and direction parameters are transmitted to the rotation interface of the camera to trigger the rotation of the camera and move it to the monitoring area of ​​interest.

[0070] As an example, first implement zoom: Camera zoom: Set the minimum detection pixel size to , then the zoom factor P is: .in, The actual monitor pixel size.

[0071] Then, judge If the value is not within the scale range of the real-time monitoring image, it means that the camera may only capture part of the area of ​​interest. If all boundary points are not within the scale range of the image, it means that the area of ​​interest is no longer within the shooting range. When a point is not within the shooting range, the camera needs to be adjusted.

[0072] Assuming the width of the imaging area is W, the height is H, and the focal length is f, the above angle calculation is as follows:

[0073] in is the vertical field of view, is the horizontal field of view.

[0074] For the points mapped in the area of ​​interest, determine whether there are any points outside the area by the following methods: if , , , If one of the four inequalities holds true, the camera needs to be moved, otherwise, it does not need to be moved.

[0075] The specific steps of camera movement are as follows: Determine the center coordinates of the mapped area. For simplicity, we can first find the circumscribed rectangle of the area. First find the four extreme points of the area.

[0076] Further find the center coordinates of the rectangle: ,

[0077] The center of the surveillance image is , the center coordinates of the region's circumscribed rectangle are ,according to and The value is used to determine the direction of camera movement: if , move to the upper left.

[0078] if , move upward.

[0079] if , move to the upper right.

[0080] if , move due right.

[0081] if , move downward and right.

[0082] if , move straight down.

[0083] if , move to the lower left.

[0084] if , move to the left.

[0085] Furthermore, the camera moves horizontally at an angle for:

[0086] The vertical angle of the camera for: .

[0087] The present invention provides an adaptive monitoring and identification method for a specified area. The method adapts the focus area in the monitoring picture, does not require manual inspection and manual adjustment, greatly saves manpower, makes monitoring efficient and feasible, and makes the method feasible for some scene monitoring applications that require real-time processing.

[0088] See also Figure 2 As shown, the present invention also discloses an adaptive monitoring and identification system for a designated area, comprising: a panoramic image acquisition module 1, a focus area coordinate extraction module 2, a feature point extraction module 3, a static feature point screening module 4, a real-time image feature point extraction module 5, a feature matching and mapping module 6, a coordinate mapping module 7 and a position identification and adjustment module 8. The panoramic picture acquisition module 1 is used to acquire the video stream of the monitoring area by adjusting the camera, and then acquire the monitoring picture, and acquire the panoramic picture of the monitoring area by splicing the monitoring pictures.

[0089] The focus area coordinate extraction module 2 is used to mark the focus area in the panoramic image using the labelme tool and obtain the coordinates of the boundary points of the focus area.

[0090] The feature point extraction module 3 is used to extract feature points from the panoramic picture using a feature point extraction algorithm.

[0091] The static feature point screening module 4 is used to draw the feature points on the panoramic image and draw the convex area where the static object is located through the labelme tool. There can be multiple such convex areas. The cross multiplication method is used to determine that the feature points in the marked convex area are static feature points, and the first feature point set is generated based on the static feature points.

[0092] The real-time image feature point extraction module 5 is used to obtain the real-time monitoring image through the camera, and extract feature points from the real-time monitoring image using a feature point extraction algorithm to generate a second feature point set.

[0093] The feature matching and mapping module 6 is used to perform feature matching on the second feature point set and the first feature point set, calculate the homography matrix, and establish a mapping relationship between the real-time monitoring picture and the panoramic picture according to the homography matrix.

[0094] The coordinate mapping module 7 is used to map the coordinates of the boundary points of the focus area to the real-time monitoring picture based on the mapping relationship, and generate the mapping coordinates of the boundary points.

[0095] The position identification and adjustment module 8 is used to determine whether the position of the camera needs to be adjusted according to the mapping coordinates of the boundary points and the range of the real-time monitoring picture. If so, the camera movement angle is determined according to the mapping coordinates of the boundary points, and the camera movement is controlled to perform video acquisition.

[0096] The specific implementation of the adaptive monitoring and identification system for a designated area of ​​this embodiment is basically the same as the specific implementation of the adaptive monitoring and identification method for a designated area described above, and will not be described in detail here.

[0097] The present invention also discloses an adaptive monitoring and identification device for a designated area, comprising a processor and a memory; wherein, when the processor executes an adaptive monitoring and identification program for a designated area stored in the memory, the steps of the adaptive monitoring and identification method for a designated area as described in any one of the above items are implemented.

[0098] Furthermore, the adaptive monitoring and identification device for a designated area in this embodiment may also include: The input interface is used to obtain the adaptive monitoring and recognition program of the specified area imported from the outside, and save the obtained adaptive monitoring and recognition program of the specified area to the memory, and can also be used to obtain various instructions and parameters transmitted by the external terminal device, and transmit them to the processor, so that the processor can use the above various instructions and parameters to carry out corresponding processing. In this embodiment, the input interface can specifically include but is not limited to a USB interface, a serial interface, a voice input interface, a fingerprint input interface, a hard disk reading interface, etc.

[0099] The output interface is used to output various data generated by the processor to the terminal device connected thereto, so that other terminal devices connected to the output interface can obtain various data generated by the processor. In this embodiment, the output interface may specifically include but is not limited to a USB interface, a serial interface, etc.

[0100] The communication unit is used to establish a remote communication connection between the adaptive monitoring and identification device in the specified area and the external server, so that the adaptive monitoring and identification device in the specified area can mount the image file to the external server. In this embodiment, the communication unit may specifically include but is not limited to a remote communication unit based on wireless communication technology or wired communication technology.

[0101] The keyboard is used to obtain various parameter data or instructions input by the user by tapping the keycaps in real time.

[0102] A display is used to display real-time information related to the adaptive monitoring and identification process running in a specified area.

[0103] The mouse can be used to assist users in inputting data and simplify user operations.

[0104] The present invention also discloses a readable storage medium, wherein the readable storage medium includes a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a hard disk, a removable hard disk, a CD-ROM or any other form of storage medium known in the technical field. The readable storage medium stores an adaptive monitoring and identification program for a specified area, and when the adaptive monitoring and identification program for a specified area is executed by a processor, the steps of the adaptive monitoring and identification method for a specified area as described in any one of the above items are implemented.

[0105] In summary, the present invention realizes comprehensive, accurate and intelligent monitoring of the monitoring area through the adaptive monitoring identification method, improves the efficiency and accuracy of the monitoring system, and enhances the flexibility and adaptability of the monitoring system.

[0106] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. As for the method disclosed in the embodiment, since it corresponds to the system disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0107] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in the above description according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0108] In the several embodiments provided by the present invention, it should be understood that the disclosed systems, systems and methods can be implemented in other ways. For example, the system embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of systems or units, which can be electrical, mechanical or other forms.

[0109] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0110] In addition, each functional module in each embodiment of the present invention may be integrated into one processing unit, or each module may exist physically separately, or two or more modules may be integrated into one unit.

[0111] Similarly, each processing unit in each embodiment of the present invention may be integrated into one functional module, or each processing unit may exist physically, or two or more processing units may be integrated into one functional module.

[0112] The steps of the method or algorithm described in conjunction with the embodiments disclosed herein may be implemented directly using hardware, a software module executed by a processor, or a combination of the two. The software module may be placed in a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

[0113] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0114] The above is a detailed introduction to the adaptive monitoring and identification method, system, device and readable storage medium for a specified area provided by the present invention. Specific examples are used herein to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. It should be pointed out that for ordinary technicians in this technical field, without departing from the principles of the present invention, several improvements and modifications can be made to the present invention, and these improvements and modifications also fall within the scope of protection of the present invention.

Claims

1. An adaptive monitoring and identification method for a specified area, characterized in that: include: By adjusting the camera to obtain the video stream of the monitoring area, and then obtaining the monitoring picture, and by splicing the monitoring pictures, obtaining the panoramic picture of the monitoring area; Use the labelme tool to mark the area of ​​interest in the panoramic image and obtain the coordinates of the boundary points of the area of ​​interest; Use feature point extraction algorithm to extract feature points from the panoramic image; Draw the feature points on the panoramic image, and use the labelme tool to draw the convex area where the static object is located. Use the cross multiplication discriminant method to extract the static feature points in the convex area, and generate the first feature point set based on the static feature points. Acquire a real-time monitoring picture through a camera, and extract feature points from the real-time monitoring picture using a feature point extraction algorithm to generate a second feature point set; Performing feature matching on the second feature point set and the first feature point set, calculating a homography matrix, and establishing a mapping relationship between the real-time monitoring image and the panoramic image according to the homography matrix; Based on the mapping relationship, the coordinates of the boundary points of the focus area are mapped to the real-time monitoring image to generate the mapping coordinates of the boundary points; According to the mapping coordinates of the boundary points and the range of the real-time monitoring picture, it is determined whether the position of the camera needs to be adjusted. If so, the movement angle of the camera is determined according to the mapping coordinates of the boundary points, and the camera movement is controlled to perform video acquisition.

2. The method for adaptive monitoring and identification of a designated area according to claim 1, characterized in that: The method of obtaining a video stream of a monitoring area by adjusting the camera, and then obtaining a monitoring picture, and obtaining a panoramic picture of the monitoring area by splicing the monitoring pictures, includes: Move the camera to its extreme position so that the camera's viewing angle can maximize the monitoring area and collect video streams; For the video stream collected at each extreme position, monitoring pictures are obtained by frame extraction, and the image stitching algorithm is applied to combine the monitoring pictures into a panoramic picture.

3. The adaptive monitoring and identification method for a designated area according to claim 1, characterized in that: The method of drawing the feature points on the panoramic image, drawing the convex area where the static object is located by using the labelme tool, extracting the static feature points in the convex area by using the cross multiplication discriminant method, and generating the first feature point set based on the static feature points includes: Use the Pillow software tool to draw the feature points on the panoramic image, use the labelme tool to frame the feature points on the static object with a convex polygon, obtain the vertex coordinates of the convex polygon, and generate a convex polygon vertex set ; Based on the convex polygon vertex set, the cross product discrimination method is used to determine whether any feature point t is a static feature point. If so, the feature point t is recorded in the first feature point set; the coordinates of the feature point t are .

4. The method for adaptive monitoring and identification of a designated area according to claim 3, characterized in that: The method of using the cross product determination method to determine whether any feature point t is a static feature point includes: Calculate the values ​​of the following discriminants respectively: If the values ​​of the discriminant are all positive or all negative, the feature point t is a static feature point.

5. The method for adaptive monitoring and identification of a designated area according to claim 1, characterized in that: The feature matching of the second feature point set with the first feature point set, calculating the homography matrix, and establishing a mapping relationship between the real-time monitoring picture and the panoramic picture according to the homography matrix includes: Based on Lowe's algorithm, feature matching is performed on the second feature point set and the first feature point set to determine four pairs of matching points; Each pair of matching points The conversion relationship is as follows: in, is the feature point in the first feature point set, and its coordinates are , is the feature point in the second feature point set, and its coordinates are , is the homography matrix; Based on the above transformation relationship and four pairs of matching points, the value of the homography matrix is ​​calculated using the Gaussian elimination method; The mapping relationship between the real-time monitoring image and the panoramic image is: in, The coordinates of the boundary points of the panoramic image focus area. It is the mapping coordinates of the boundary points in the monitoring image.

6. The method for adaptive monitoring and identification of a designated area according to claim 5, characterized in that: The step of mapping the boundary point coordinates of the focus area to the real-time monitoring image based on the mapping relationship to generate the mapping coordinates of the boundary points includes: The mapping coordinates of the boundary points are calculated according to the following formula: 。 7. The method for adaptive monitoring and identification of a designated area according to claim 6, characterized in that: The determining whether the position of the camera needs to be adjusted according to the mapping coordinates of the boundary points and the range of the real-time image, and if so, determining the moving angle of the camera according to the mapping coordinates of the boundary points, includes: Set the minimum detection pixel size of the camera to , then the zoom factor P is: ; in, To monitor the actual pixel size; judge , , , Whether one of the four inequalities holds true; If so, you need to adjust the position of the camera; Based on the mapping coordinates of the boundary points, the four extreme points of the circumscribed rectangle of the corresponding area are calculated using the following formula: The center coordinates of the circumscribed rectangle are calculated using the following formula: , ； According to the center coordinates of the monitoring image and the center coordinates of the bounding rectangle The relative position of determines the moving direction of the camera; The horizontal and vertical movement angles of the camera are calculated using the following formula: in, is the horizontal movement angle of the camera, is the vertical movement angle of the camera.

8. An adaptive monitoring and identification system for a designated area, characterized in that: The system adopts the adaptive monitoring and identification method of the designated area as claimed in any one of claims 1 to 7; The system comprises: A panoramic picture acquisition module is used to acquire the video stream of the monitoring area by adjusting the camera, and then acquire the monitoring picture, and acquire the panoramic picture of the monitoring area by splicing the monitoring pictures; The module for extracting coordinates of the region of interest is used to mark the region of interest in the panoramic image using the labelme tool and obtain the coordinates of the boundary points of the region of interest; A feature point extraction module is used to extract feature points from the panoramic image using a feature point extraction algorithm; The static feature point screening module is used to draw the feature points on the panoramic image, draw the convex area where the static object is located through the labelme tool, use the cross multiplication discriminant method to extract the static feature points in the convex area, and generate the first feature point set based on the static feature points; A real-time image feature point extraction module is used to obtain a real-time monitoring image through a camera, and extract feature points from the real-time monitoring image using a feature point extraction algorithm to generate a second feature point set; A feature matching and mapping module, used to perform feature matching on the second feature point set and the first feature point set, calculate a homography matrix, and establish a mapping relationship between the real-time monitoring image and the panoramic image according to the homography matrix; A coordinate mapping module is used to map the coordinates of the boundary points of the focus area to the real-time monitoring picture based on the mapping relationship to generate the mapping coordinates of the boundary points; The position recognition and adjustment module is used to determine whether the position of the camera needs to be adjusted based on the mapping coordinates of the boundary points and the range of the real-time monitoring image. If so, the camera movement angle is determined based on the mapping coordinates of the boundary points, and the camera movement is controlled to perform video acquisition.

9. An adaptive monitoring and identification device for a designated area, characterized in that: include: A memory, used for storing an adaptive monitoring and recognition program for a designated area; A processor is used to implement the steps of the adaptive monitoring and identification method for a designated area as described in any one of claims 1 to 7 when executing the adaptive monitoring and identification program for the designated area.

10. A readable storage medium, characterized in that: The readable storage medium stores an adaptive monitoring and identification program for a specified area, and when the adaptive monitoring and identification program for a specified area is executed by a processor, the steps of the adaptive monitoring and identification method for a specified area as described in any one of claims 1 to 7 are implemented.

Images