Method and apparatus for monitoring adjustment of a device

Abstract

The application discloses a kind of adjusting method and device of monitoring equipment, in which, according to the multiple images collected by monitoring equipment, the attribute graph corresponding to the multiple images is determined respectively, and the evaluation index is obtained by evaluating attribute graph, and finally the adjustment information of monitoring equipment is determined based on evaluation index, the adjustment information can be used as the reference information for adjusting monitoring equipment, so this scheme can realize the dynamic adjustment of monitoring equipment, and improves the monitoring ability of monitoring equipment.

Description

Technical Field

[0001] This application relates to the field of computers, and more particularly to a method and apparatus for adjusting a monitoring device. Background Technology

[0002] With the development of communication technology and the widespread application of computer vision, more and more monitoring devices need to monitor different types of targets and ensure the image quality of the monitored targets, such as resolution, sharpness, brightness, and occlusion. However, some monitoring devices often do not take these aspects into account when they are deployed, resulting in low availability of the monitoring equipment. Summary of the Invention

[0003] This application provides a method and apparatus for adjusting monitoring equipment to improve the monitoring capabilities of the monitoring equipment.

[0004] In a first aspect, embodiments of this application provide a method for adjusting a monitoring device. The method includes: acquiring multiple images collected by the monitoring device within a target time period, wherein each of the multiple images contains a target object, and the target objects contained in the multiple images correspond to the same target type; determining at least one attribute map based on the multiple images and at least one attribute corresponding to the target type, wherein the at least one attribute map corresponds one-to-one with at least one attribute; evaluating the at least one attribute map to obtain at least one evaluation index corresponding to the monitoring device; and adjusting the monitoring device based on the at least one evaluation index.

[0005] The above scheme determines the attribute maps corresponding to the multiple images collected by the monitoring equipment, evaluates the attribute maps to obtain evaluation indicators, and finally determines the adjustment information of the monitoring equipment based on the evaluation indicators. This adjustment information can be used as reference information for adjusting the monitoring equipment. Therefore, this scheme can realize the dynamic adjustment of the monitoring equipment and improve the monitoring capabilities of the monitoring equipment.

[0006] In one possible implementation, determining at least one attribute map based on multiple images and at least one attribute corresponding to the target type includes: identifying multiple attribute features of the target object and the position of each attribute feature in the images from the multiple images respectively; for each attribute feature, integrating the attribute feature of the target object in the multiple images according to the time sequence of the acquisition time of the multiple images and the position of the attribute feature in the images to form an attribute map corresponding to the attribute feature.

[0007] The above solution has two advantages. First, the attribute map contains information from multiple images over a period of time, which can more accurately reflect the actual monitoring information of the monitoring equipment. Second, constructing different attribute maps allows for the evaluation of the monitoring equipment from various aspects, resulting in a more comprehensive evaluation. Adjustments to the monitoring equipment based on these evaluations can then enhance its monitoring capabilities.

[0008] In one possible implementation, adjusting the monitoring device based on at least one evaluation indicator includes: obtaining the priority of the output evaluation indicator; the priority is related to the business to be performed by the monitoring device; and adjusting the monitoring device based on the indicator value of the output evaluation indicator and the priority of the evaluation indicator.

[0009] The above scheme links evaluation indicators with the business that the monitoring equipment needs to perform, and prioritizes adjustments to the monitoring equipment based on the business that the monitoring equipment needs to perform. This helps to accurately determine the adjustment information used to adjust the monitoring equipment, and can ensure that the monitoring equipment is in a better monitoring state as much as possible, thereby improving the monitoring capabilities of the monitoring equipment.

[0010] In one possible implementation method, the monitoring equipment is adjusted based on the output evaluation index values ​​and the priority of the evaluation indexes, including: determining adjustment suggestions for the monitoring parameters of the monitoring equipment based on the output evaluation index values ​​and the priority of the evaluation indexes; and sending the adjustment suggestions to the operation and maintenance personnel so that the operation and maintenance personnel can adjust the monitoring equipment.

[0011] The above solution allows maintenance personnel to adjust the monitoring equipment according to the adjustment suggestions, which can accurately adjust the monitoring equipment and thus improve its monitoring capabilities.

[0012] In one possible implementation, the monitoring parameters of the monitoring equipment are adjusted based on the output evaluation index values ​​and the priority of the evaluation indexes. This includes: when the evaluation index is related to other evaluation indexes, determining the adjustment information of the monitoring parameters of the monitoring equipment based on the evaluation index and the priority of the related other evaluation indexes; wherein the priority is positively correlated with the importance of the business processed by the monitoring equipment as reflected by the evaluation index; and adjusting the monitoring parameters of the monitoring equipment based on the determined adjustment information.

[0013] The above scheme, by adjusting the monitoring parameters of the monitoring equipment according to the evaluation indicators and other related evaluation indicators, helps to accurately determine the adjustment information for adjusting the monitoring equipment, enabling a more comprehensive adjustment of the monitoring equipment, putting the monitoring equipment in a better monitoring state, and improving the monitoring capabilities of the monitoring equipment.

[0014] In one possible implementation method, adjusting the monitoring parameters of the monitoring device based on the determined adjustment information includes: adjusting the monitoring parameters of the monitoring device according to the adjustment information so that the highest priority evaluation indicator reaches the optimal level, and other evaluation indicators other than the highest priority evaluation indicator meet the preset requirements.

[0015] The above solution uses optimized monitoring parameters to adjust the monitoring equipment, which helps to accurately determine the adjustment information used to adjust the monitoring equipment, enabling the monitoring equipment to be in a better monitoring state and improving the monitoring capabilities of the monitoring equipment.

[0016] In one possible implementation method, the evaluation indicators include at least one of the following: image quality evaluation, scene coverage evaluation, scene applicability evaluation, or target applicability evaluation.

[0017] The above-mentioned solution, through image quality evaluation, scene coverage evaluation, scene applicability evaluation, and target applicability evaluation, can comprehensively and accurately analyze the images monitored by the monitoring equipment, which helps to accurately determine the adjustment information used to adjust the monitoring equipment.

[0018] In one possible implementation method, the image quality evaluation specifically includes at least one of the following: the side angle size, pitch angle size, rotation angle size, resolution size, light intensity, or target blur of the monitoring device; the scene coverage evaluation specifically includes: the confidence level of the scene coverage degree of the monitoring device at different angles; the scene applicability evaluation specifically includes at least one of the following: the screen occlusion of the monitoring device, background applicability, or site applicability; the target applicability evaluation specifically includes at least one of the following: the target type ratio analysis, the main traffic flow direction analysis, or the capture position analysis of the monitoring device.

[0019] The above scheme includes evaluation indicators for image quality assessment, which can accurately evaluate the quality of various types of targets captured by the monitoring equipment; evaluation indicators for scene coverage assessment, which can accurately evaluate whether there are blind spots or invalid areas in the image monitored by the monitoring equipment; evaluation indicators for scene suitability assessment, which can accurately evaluate the suitability of the image monitored by the monitoring equipment for computer vision tasks; and evaluation indicators for target suitability assessment, which can accurately evaluate the proportion of targets captured by the monitoring equipment, the angle between various types of targets and the monitoring equipment, and the position of various types of targets in the monitoring image. This helps to accurately determine the adjustment information used to adjust the monitoring equipment.

[0020] Secondly, embodiments of this application provide an adjustment device for a monitoring device, including an acquisition unit, a determination unit, an evaluation unit, and an adjustment unit. The acquisition unit is used to acquire multiple images collected by the monitoring device within a target time period, each image containing a target object, and the target objects contained in each image corresponding to the same target type. The determination unit is used to determine at least one attribute map based on the multiple images and at least one attribute corresponding to the target type, wherein the at least one attribute map corresponds one-to-one with at least one attribute. The evaluation unit is used to evaluate the at least one attribute map to obtain at least one evaluation index corresponding to the monitoring device. The adjustment unit is used to adjust the monitoring device according to the at least one evaluation index.

[0021] In one possible implementation, the determining unit is used to: identify multiple attribute features of the target object and the corresponding position of each attribute feature in the images from multiple images respectively; for each attribute feature, according to the time sequence of the acquisition time of the multiple images and the corresponding position of the attribute feature in the images, integrate the attribute feature of the target object in the multiple images to form an attribute map corresponding to the attribute feature.

[0022] In one possible implementation, the adjustment unit is used to: obtain the priority of the output evaluation index; the priority is related to the business to be performed by the monitoring device; and adjust the monitoring device according to the index value of the output evaluation index and the priority of the evaluation index.

[0023] In one possible implementation, the adjustment unit is used to: adjust the monitoring equipment based on the output evaluation index values ​​and the priority of the evaluation indexes, including: determining adjustment suggestions for the monitoring parameters of the monitoring equipment based on the output evaluation index values ​​and the priority of the evaluation indexes; and sending the adjustment suggestions to the operation and maintenance personnel so that the operation and maintenance personnel can adjust the monitoring equipment.

[0024] In one possible implementation, the adjustment unit is used to: determine the adjustment information of the monitoring parameters of the monitoring device based on the priority of the evaluation indicator and the other related evaluation indicators when the evaluation indicator is related to other evaluation indicators; wherein the priority is positively correlated with the importance of the business processed by the monitoring device reflected by the evaluation indicator; and adjust the monitoring parameters of the monitoring device according to the determined adjustment information.

[0025] In one possible implementation, the adjustment unit is used to: adjust the monitoring parameters of the monitoring equipment according to the adjustment information, so that the highest priority evaluation index reaches the optimal level, and other evaluation indicators other than the highest priority evaluation index meet the preset requirements.

[0026] In one possible implementation method, the evaluation indicators include at least one of the following: image quality evaluation, scene coverage evaluation, scene applicability evaluation, or target applicability evaluation.

[0027] In one possible implementation method, the image quality evaluation specifically includes at least one of the following: the side angle size, pitch angle size, rotation angle size, resolution size, light intensity, or target blur of the monitoring device.

[0028] In one possible implementation method, scene coverage evaluation specifically includes: confidence level of scene coverage of monitoring equipment from different angles.

[0029] In one possible implementation method, the scene suitability evaluation specifically includes at least one of the following: the screen occlusion of the monitoring equipment, background suitability, or site suitability.

[0030] In one possible implementation method, the target suitability evaluation specifically includes at least one of the following: target type ratio analysis, main traffic flow direction analysis, or capture location analysis of the monitoring equipment.

[0031] Thirdly, embodiments of this application also provide a computer-readable storage medium storing computer-readable instructions, which, when read and executed by a computer, implement any of the methods described in the first aspect. Attached Figure Description

[0032] Figure 1 A flowchart illustrating a method for adjusting a monitoring device provided in an embodiment of this application;

[0033] Figure 2 This is a schematic diagram of the structure of an adjustment device for a monitoring device provided in an embodiment of this application;

[0034] Figure 3 This is a schematic diagram of the structure of an adjustment device for a monitoring device provided in an embodiment of this application. Detailed Implementation

[0035] Figure 1 This is a flowchart illustrating a method for adjusting a monitoring device according to an embodiment of this application. The method can be executed by an adjustment device for the monitoring device, which can be a terminal device, a server, an application program for the terminal device, or an application program for the server. This application does not limit the entity executing this method.

[0036] The method includes the following steps:

[0037] Step 101: Acquire multiple images collected by the monitoring equipment within the target time period.

[0038] The plurality of images each contain a target object, and the targets contained in the plurality of images correspond to the same target type.

[0039] In one possible implementation, the target type includes one or more of the following: face, body, motor vehicle, and non-motor vehicle.

[0040] The following specific example illustrates how to acquire multiple images collected by a monitoring device within a target time period.

[0041] The monitoring equipment is pre-set to detect faces. In a video feed, at time t1, a pedestrian appears in the monitored frame. Face detection technology is used to identify a face image, namely face image A, and the position 'a' of face image A in the monitored frame. This position is also called the face marker 'a', which can be the center point of face image A or any pixel in face image A; this application does not limit this. At time t2, another pedestrian appears in the monitored frame. Face detection technology is used to identify another face image, namely face image B, and the position 'b' of face image B in the monitored frame. This position is also called the face marker 'b'. The monitored frames at time t1 and time t2 can be video frames from adjacent times or video frames from relatively distant times; this application does not limit this. This application does not limit the technical means used for face detection in the monitored frame.

[0042] Step 102: Determine at least one attribute map based on multiple images and at least one attribute corresponding to the target type.

[0043] In this context, at least one attribute map corresponds one-to-one with at least one attribute, meaning that an attribute map is composed of the same attribute of the target type corresponding to multiple images.

[0044] In one possible implementation, the dimensions of the attribute graph are the same as those of the screen monitored by the monitoring device.

[0045] In one possible implementation, the attributes of a human face include one or more of angle, resolution, clarity, occlusion, etc.; the attributes of a human body include one or more of angle, resolution, clarity, occlusion, clothing, etc.; the attributes of a motor vehicle include one or more of angle, resolution, clarity, occlusion, license plate, etc.; and the attributes of a non-motorized vehicle include one or more of angle, resolution, clarity, occlusion, whether there are people on the motor vehicle, etc.

[0046] One possible implementation involves identifying multiple attribute features of a target object and the corresponding position of each attribute feature in the images from multiple images. For each attribute feature, according to the temporal order of the images' acquisition times and its corresponding position in the images, the attribute feature of the target object in the multiple images is integrated to form an attribute map corresponding to that attribute feature. The position of each attribute feature in the image corresponds to its position in the monitoring screen based on the image acquired according to the target type. This approach has two advantages: firstly, the attribute map contains information from multiple images over a period of time, enabling a more accurate reflection of the actual monitoring information from the monitoring device; secondly, constructing different attribute maps allows for evaluation of the monitoring device from various aspects, resulting in a more comprehensive evaluation. Adjustments to the monitoring device based on these evaluation results can then be made to improve its monitoring capabilities.

[0047] The following section will use the target type of human face to illustrate how to determine an attribute graph.

[0048] In step 101, face images A and B are used to obtain multiple facial attributes through certain technical means. These techniques can include angle detection, occlusion detection, etc., from machine learning, and this application does not limit the specific techniques used. These facial attributes include one or more of angle, resolution, clarity, and occlusion. This application uses the angle attribute as an example to illustrate the angle attribute map. For example, the angle attribute obtained from face image A through angle detection and other techniques is 0, meaning the face in the captured face image A is frontal; the angle attribute obtained from face image B through angle detection and other techniques is +30, meaning the face in the captured face image B is shifted 30 degrees to the right. The positive or negative value of the angle attribute represents the direction of shift, and the absolute value represents the degree of shift.

[0049] Since the dimensions of the attribute map are the same as those of the monitored image, the corresponding position of a face image in the attribute map can be determined based on the temporal sequence of multiple image acquisitions and the face markers. That is, face marker 'a' in face image A corresponds to position 'a' in the angular attribute map, and face marker 'b' in face image B corresponds to position 'b' in the angular attribute map. Therefore, position 'a' in the angular attribute map corresponds to 0, and position 'b' corresponds to 30. If multiple face markers are located at the same position in the attribute map, the average of all attributes at that position is taken as the attribute for that position. If no face marker corresponds to a position in the attribute map, that position can be set to a fixed value, such as 255.

[0050] Step 103: Evaluate at least one attribute graph to obtain at least one evaluation index corresponding to the monitoring device.

[0051] In one possible implementation method, the evaluation indicators include at least one of the following: image quality evaluation, scene coverage evaluation, scene applicability evaluation, or target applicability evaluation. This solution, through image quality evaluation, scene coverage evaluation, scene applicability evaluation, and target applicability evaluation, can comprehensively and accurately analyze the images monitored by the surveillance equipment, which helps to accurately determine the adjustment information used to adjust the surveillance equipment.

[0052] In one possible implementation method, the image quality evaluation specifically includes at least one of the following: the side angle size, pitch angle size, rotation angle size, resolution size, light intensity, or target blur of the monitoring device.

[0053] In one possible implementation method, scene coverage evaluation specifically includes: confidence level of scene coverage of monitoring equipment from different angles.

[0054] In one possible implementation method, the scene suitability evaluation specifically includes at least one of the following: the screen occlusion of the monitoring equipment, background suitability, or site suitability.

[0055] In one possible implementation method, the target suitability evaluation specifically includes at least one of the following: target type ratio analysis, main traffic flow direction analysis, or capture location analysis of the monitoring equipment.

[0056] The above scheme includes the following evaluation indicators: Image quality evaluation, which accurately assesses the quality of various types of targets captured by the monitoring equipment; Scene coverage evaluation, which accurately assesses whether the monitored image contains blind spots or invalid areas; Scene suitability evaluation, which accurately assesses the suitability of the monitored image for computer vision tasks; and Target suitability evaluation, which accurately evaluates the proportion of targets captured by the monitoring equipment, the angle between various types of targets and the monitoring equipment, and the position of various types of targets in the monitored image. This helps to accurately determine the adjustment information used to adjust the monitoring equipment.

[0057] Step 104: Adjust the monitoring equipment according to at least one evaluation indicator.

[0058] The above scheme determines the attribute maps corresponding to the multiple images collected by the monitoring equipment, evaluates the attribute maps to obtain evaluation indicators, and finally determines the adjustment information of the monitoring equipment based on the evaluation indicators. This adjustment information can be used as reference information for adjusting the monitoring equipment. Therefore, this scheme can realize the dynamic adjustment of the monitoring equipment and improve the monitoring capabilities of the monitoring equipment.

[0059] In one possible implementation, at least one attribute map is evaluated. The attribute map can be input into a trained neural network model corresponding to the evaluation index. This application does not limit this method.

[0060] The following sections will explain the evaluation methods for image quality, scene coverage, scene applicability, and target applicability.

[0061] Method 1: Image quality evaluation.

[0062] Image quality evaluation primarily considers the quality of the captured target, such as whether the side angle is too large, the top angle is too large, the resolution is too low, the captured target is too dark or too bright, or the target motion is blurred. Among these, an excessive side angle may be manifested as a significant side angle between the monitoring equipment and the main direction of pedestrian flow; an excessive top angle may be manifested as the monitoring equipment only being able to capture a portion of the target; and low resolution may be manifested as the captured target being too large or the target being concentrated at the far end of the monitoring screen.

[0063] In one possible implementation method, the evaluation indicators corresponding to the image quality assessment include at least one of the following: the side angle, pitch angle, rotation angle, resolution, light intensity, or target blur of the monitoring equipment and the captured target. The output indicators are all absolute values, with positive and negative values ​​representing the direction of offset, and the absolute value representing the magnitude or degree of offset. For example, the value range for the side angle is -180 to +180.

[0064] The above solution can accurately evaluate the quality of various types of targets captured by the monitoring equipment, which helps to accurately determine the adjustment information used to adjust the monitoring equipment.

[0065] Method 2: Scenario Coverage Evaluation.

[0066] In one possible implementation, scene coverage evaluation mainly considers whether the monitoring screen is too small, resulting in blind spots in the captured road, or whether the screen contains inaccessible areas such as pools of water. For example, if the target is a human body, but only part of the body is shown, the monitoring device may be monitoring an area that is too small. Similarly, if the target only appears at the edge of the monitoring screen, the screen may contain inaccessible areas such as pools of water.

[0067] In one possible implementation method, the evaluation index corresponding to the scene coverage evaluation includes the confidence level of the scene coverage of the monitoring device from different angles. For example, the confidence levels in the four directions of up, down, left, and right respectively represent whether there are blind spots or invalid areas at the four edges. Positive and negative values ​​are used to distinguish whether to zoom in or zoom out. The absolute value represents the confidence strength, that is, the degree of adjustment required. The larger the value, the more adjustments are needed.

[0068] The above solution can accurately evaluate whether there are blind spots or invalid areas in the image monitored by the monitoring equipment, which helps to accurately determine the adjustment information for adjusting the monitoring equipment.

[0069] Method 3: Scenario Applicability Evaluation.

[0070] In one possible implementation, scene suitability evaluation mainly considers the applicability of the monitoring screen content and site characteristics to the computer vision task. For example, if the monitoring screen is found to be severely obstructed by leaves or other objects, the monitoring equipment is clearly unsuitable, and this problem cannot be solved by adjusting camera parameters. Another example is a location primarily targeting motor vehicles, where the monitoring equipment is deployed next to the sidewalk, constantly capturing pedestrians.

[0071] In one possible implementation method, the evaluation index corresponding to scene applicability evaluation includes at least one of the following: screen occlusion of the monitoring equipment, background applicability, or site applicability. Screen occlusion mainly refers to the monitoring equipment being obstructed by certain objects, including leaves, parasols, vehicles, etc., making the monitoring image unusable. Background applicability mainly refers to the fact that the background of the scene captured by the monitoring equipment is very complex, such as brightly colored carpets, walls, etc., making it easy for the colors of people's clothing to be confused with the background. Site applicability refers to the fact that the place monitored by the monitoring equipment may not be suitable for the corresponding computer vision task. For example, the computer vision task considers people wearing the same clothes to be the same person, but all the people in the place monitored by the monitoring equipment are wearing uniform work clothes; or the computer vision task is to analyze license plates, but the monitoring equipment is capturing a sidewalk. The three parts of screen occlusion, background applicability, and site applicability are all confidence outputs, with a confidence value range of [0,1]. The higher the confidence, the worse the applicability.

[0072] The above solution can accurately evaluate the suitability of the images monitored by the monitoring equipment for computer vision tasks, which helps to accurately determine the adjustment information used to adjust the monitoring equipment.

[0073] Method 4: Evaluation of the applicability of the objectives.

[0074] In one possible implementation method, the target applicability evaluation mainly considers that various types of targets can be captured with high quality. For example, if a pedestrian is walking towards the monitoring device, the face and body can be captured very well. However, if the monitoring device is in the opposite direction to the flow of people, it will result in the inability to capture faces.

[0075] In one possible implementation method, the evaluation indicators corresponding to the target suitability evaluation include at least one of the following: target type ratio analysis, main traffic flow direction analysis, or capture location analysis.

[0076] Target type ratio analysis: For a certain computer vision task, such as human body recognition, if the monitoring equipment monitors a large number of motor vehicles and non-motor vehicles, and the human body parts of people riding electric bikes and passengers are almost unusable due to occlusion, then the human body recognition algorithm is unusable.

[0077] One possible implementation involves statistically analyzing the proportions of various targets (such as faces, bodies, non-motorized vehicles, and motorized vehicles) in the monitored area. Based on these proportions, the number of people walking, cycling, or riding in vehicles within the monitored area is determined. Finally, the applicability of the monitoring equipment's feed to the computer vision task for that area is assessed based on the number of people walking, cycling, or riding in vehicles. The target type proportion analysis output is the proportion of each target: faces, bodies, non-motorized vehicles, and motorized vehicles.

[0078] The analysis of the direction of pedestrian traffic mainly considers whether the shooting angle of the surveillance equipment is opposite to the direction of pedestrian traffic. For example, if the shooting angle of the surveillance equipment is opposite to the direction of pedestrian traffic, the surveillance equipment can basically only capture back views.

[0079] One possible implementation involves statistically analyzing the proportions of the front, side, and back views of the captured target to infer the direction of pedestrian traffic and the angle of the monitoring equipment. The output of the pedestrian traffic direction analysis is the statistically analyzed proportions of the front, side, and back views.

[0080] Capture location analysis mainly considers the specific location of the captured target. Based on prior information, the quality is often best when the target is close to the lower edge of the monitoring screen. Therefore, capture location analysis mainly analyzes whether the targets captured by the monitoring equipment are mainly concentrated at the lower edge of the captured image. If not, the monitoring equipment may be deployed in an unreasonable manner.

[0081] One possible implementation involves calculating the distance between the captured target location and the bottom edge of the image captured by the monitoring device. The targets are then sorted by distance from smallest to largest, and the cumulative percentage of captured targets and their average distance are listed separately. For example, if the average distance is 30 pixels, 10% of motor vehicle targets are captured; if the average distance is 45 pixels, 20% of motor vehicle targets are captured. Different target types are counted independently. The capture location analysis output is the average distance of each target type from the bottom edge of the captured image when each type of target accounts for 50% of the total.

[0082] The above solution can accurately evaluate the proportion of targets captured by the monitoring equipment, the angle between each type of target and the monitoring equipment, and the position of each type of target in the monitoring screen, which helps to accurately determine the adjustment information used to adjust the monitoring equipment.

[0083] In one possible implementation, after step 105 above, the adjustment information is sent to the monitoring device, which then makes corresponding adjustments based on the adjustment information. For example, adjusting the angle, focal length, etc.

[0084] In another possible implementation, the adjustment information determined in step 105 above includes first adjustment information and second adjustment information.

[0085] The adjustment specified in the first adjustment information is one that the monitoring equipment can automatically complete. Therefore, the first adjustment information can be sent to the monitoring equipment, which will then make corresponding adjustments based on it. The first adjustment information may include, for example, angle and focal length.

[0086] One possible implementation involves obtaining the priority of the output evaluation metrics. This priority is related to the business that the monitoring equipment needs to perform. Based on the metric values ​​and priorities of the output evaluation metrics, the monitoring equipment is adjusted. For example, if the monitoring equipment needs to perform face detection, it needs to ensure that the captured faces are frontal, meaning the ideal face angle range is (-30°, 30°). Therefore, angle has the highest priority among the evaluation metrics. During the self-adjustment process, priority is given to adjusting the monitoring equipment's angle to the optimal level, ensuring that the captured faces are frontal. Secondly, the focal length can be adjusted to ensure the clarity of the captured faces. This solution associates the evaluation metrics with the business that the monitoring equipment needs to perform, prioritizing adjustments based on the business. This helps to accurately determine the adjustment information used to adjust the monitoring equipment, ensuring it is in a better monitoring state and improving its monitoring capabilities.

[0087] In one possible implementation, when other evaluation indicators are related to the evaluation indicator, the adjustment information of the monitoring parameters of the monitoring equipment is determined based on the priority of the evaluation indicator and its related indicators. The priority is positively correlated with the importance of the business handled by the monitoring equipment as reflected by the evaluation indicator. The monitoring parameters of the monitoring equipment are then adjusted based on the determined adjustment information. In other words, adjusting the monitoring parameters of the monitoring equipment according to the priority of the evaluation indicator maintains a dynamic balance between the evaluation indicator and its related indicators, thereby optimizing the adjustment information of the monitoring parameters determined based on the evaluation indicator and its related indicators. This scheme, by adjusting the monitoring parameters of the monitoring equipment according to the evaluation indicator and its related indicators, helps to accurately determine the adjustment information used to adjust the monitoring equipment, enabling a more comprehensive adjustment of the monitoring equipment, placing it in a better monitoring state, and improving the monitoring capabilities of the monitoring equipment.

[0088] The second adjustment information indicates adjustments that the monitoring equipment cannot make automatically and require manual adjustment. This second adjustment information can be sent to the monitoring center or to a handheld terminal device used by maintenance personnel. The maintenance personnel then manually adjust the monitoring equipment based on this second adjustment information. This second adjustment information may include, for example, analysis of screen obstruction, background suitability, site suitability, target type ratio, pedestrian traffic direction, or snapshot location.

[0089] In one possible implementation, based on the output evaluation index values ​​and their priorities, adjustment suggestions for the monitoring parameters of the monitoring equipment are determined. These suggestions are then sent to maintenance personnel for adjustment. For example, if the monitoring equipment needs to perform human body recognition, but the monitored area contains numerous motor vehicles and non-motor vehicles, and the human bodies of cyclists and passengers are almost unusable due to obstruction, then the location where the monitoring equipment is deployed is unsuitable for this service. Adjustment suggestions need to be sent to maintenance personnel, who will then redeploy the monitoring equipment according to the required service type and the suggested adjustments.

[0090] The above solution allows maintenance personnel to adjust the monitoring equipment according to the adjustment suggestions, which can accurately adjust the monitoring equipment and thus improve its monitoring capabilities.

[0091] The focal length and angle in the first adjustment information are explained below.

[0092] I. Automatic adjustment of the focal length of monitoring equipment

[0093] In one possible implementation method, the monitoring parameters of the monitoring equipment are adjusted according to the determined adjustment information, including: adjusting the monitoring parameters of the monitoring equipment according to the adjustment information so that the highest priority evaluation index reaches the optimal level, and other evaluation indicators other than the highest priority evaluation index meet the preset requirements.

[0094] For example, the primary monitoring function of the surveillance equipment is face detection. The highest priority evaluation metric related to this function is scene coverage evaluation, and other related evaluation metrics include image quality evaluation. That is, when determining the adjustment information for the surveillance equipment's monitoring parameters, the scene coverage metric is given priority. When adjusting the surveillance equipment's parameters based on scene coverage, the image quality metric must also be considered to optimize both scene coverage and image quality, ensuring that the overall determined monitoring parameters of the surveillance equipment are optimal.

[0095] In one possible implementation method, the adjustment information of the monitoring parameters of the monitoring equipment is first determined according to the highest priority evaluation index; then the adjustment information of the monitoring parameters of the monitoring equipment is determined according to other evaluation indexes. When there is a conflict between the adjustment information of the monitoring parameters of the monitoring equipment determined according to other evaluation indexes and the adjustment information of the monitoring parameters of the monitoring equipment determined according to the highest priority evaluation index, the adjustment information of the monitoring parameters of the monitoring equipment determined according to the highest priority evaluation index is fine-tuned so that the determined monitoring parameters of the monitoring equipment reach the optimal level.

[0096] The following section will explain in detail how to adjust the monitoring parameters of a monitoring device using the example of automatic focal length adjustment. Automatic focal length adjustment of a monitoring device needs to take into account scene coverage and target resolution in image quality evaluation.

[0097] In one possible implementation, the evaluation metric for scene coverage is that if the image monitored by the monitoring device is too large and the focal length needs to be reduced, the monitoring device can automatically shorten the focal length according to the determined adjustment information.

[0098] In one possible implementation method, the evaluation index for scene coverage is that if the monitored image is too small and the focal length needs to be increased, then it is necessary to balance scene coverage and target resolution, and gradually adjust according to the preset magnification resolution step size until the value of Formula 1 reaches the optimal value, i.e., the maximum value:

[0099] ∑ 目标类型 (Target applicability * Dependency strength - Scenario confidence * Scenario strength) Formula 1

[0100] In Formula 1, target suitability * dependency strength represents target importance, that is, the importance of different targets in the monitoring screen of the monitoring device. The larger the value, the more important the target is, and the more its quality needs to be guaranteed. Scene confidence * scene strength represents the degree of target adjustment, that is, the degree to which different targets in the monitoring screen of the monitoring device need to be adjusted. The larger the value, the more adjustment is needed. ∑ 目标类型 It is a summation function that sums the target values ​​corresponding to different target types. The target value is the difference between the target importance and the target adjustment degree.

[0101] In Formula 1, target suitability refers to the degree to which the capture resolution of a target type is suitable for computer vision tasks. Taking faces as an example, assuming no other influencing factors are flawed, face regions below 30 pixels may be almost unusable, face regions between 30 and 100 pixels show gradually better suitability for computer vision tasks, and face regions above 100 pixels show stable suitability. The target suitability for each target type requires engineers to pre-define the corresponding relationship function based on the characteristics of their algorithms and the requirements of the computer vision task. Generally, the trend is that extremely low resolution results in extremely low suitability, even zero; intermediate resolutions show suitability that gradually increases positively correlated with resolution; and high resolutions show relatively high suitability with little change.

[0102] In Formula 1, the dependence intensity refers to the degree of dependence of the monitoring equipment on various types of targets. Before step 101, it is necessary to set the main tasks of the monitoring equipment, such as license plate analysis, face detection, human body recognition, etc., and set the target types that the corresponding tasks mainly depend on, such as faces, human bodies, motor vehicles, etc., and set the dependence intensity of different types of targets. The higher the intensity, the more necessary it is to ensure the quality of such targets.

[0103] In Formula 1, scene confidence is used to represent the evaluation index of scene coverage, including confidence in the four directions of up, down, left, and right. Since the focal length of the monitoring equipment can only enlarge or reduce the image proportionally in length and width, the scene confidence is the average of the confidence in the four directions of scene coverage.

[0104] In Formula 1, scene intensity is used to represent the balance between scene coverage and target size. The larger the value of scene intensity, the more necessary it is to ensure scene coverage and avoid monitoring blind spots.

[0105] II. Automatic adjustment of the angle of the monitoring equipment.

[0106] Automatic adjustment of the monitoring equipment angle needs to consider the side angle, pitch angle, and rotation angle in image quality evaluation. Since the installation of monitoring equipment is determined based on the overall layout, the angle adjustment needs to be limited, that is, the maximum adjustment amount of the side angle, pitch angle, and rotation angle should be preset to prevent disruption of the overall layout.

[0107] In one possible implementation, if the monitoring equipment cannot resolve the angle issue through automatic adjustment, an adjustment prompt message needs to be sent to the on-site maintenance personnel, who will then adjust the monitoring equipment according to the prompt message.

[0108] In one possible implementation, image quality evaluation, scene coverage evaluation, and scene suitability evaluation can evaluate not only the attribute map but also the scene map plus the attribute map, thus obtaining the corresponding evaluation indicators for the monitoring equipment. A scene map refers to a single image captured by the monitoring equipment without a target, or multiple images with different targets present.

[0109] In one possible implementation, the target suitability evaluation can be performed directly using images captured by monitoring equipment.

[0110] Based on the same technological concept Figure 2 An exemplary embodiment of an adjustment device 200 for a monitoring device provided in this application is shown. For example... Figure 2 As shown, the system includes: an acquisition unit 201, a determination unit 202, an evaluation unit 203, and an adjustment unit 204. The acquisition unit 201 is used to acquire multiple images collected by the monitoring equipment within a target time period. Each of the multiple images contains a target object, and the target objects contained in each of the multiple images correspond to the same target type. The determination unit 202 is used to determine at least one attribute map based on the multiple images and at least one attribute corresponding to the target type, wherein the at least one attribute map corresponds one-to-one with at least one attribute. The evaluation unit 203 is used to evaluate the at least one attribute map to obtain at least one evaluation index corresponding to the monitoring equipment. The adjustment unit 204 is used to adjust the monitoring equipment based on the at least one evaluation index.

[0111] In one possible implementation, the determining unit 202 is used to: identify multiple attribute features of the target object and the corresponding position of each attribute feature in the images from multiple images respectively; for each attribute feature, according to the time sequence of the acquisition time of the multiple images and the corresponding position of the attribute feature in the images, integrate the attribute feature of the target object in the multiple images to form an attribute map corresponding to the attribute feature.

[0112] In one possible implementation, the adjustment unit 204 is used to: obtain the priority of the output evaluation index; the priority is related to the business to be performed by the monitoring device; and adjust the monitoring device according to the index value of the output evaluation index and the priority of the evaluation index.

[0113] In one possible implementation, the adjustment unit 204 is used to: adjust the monitoring equipment according to the index values ​​of the output evaluation indicators and the priority of the evaluation indicators, including: determining adjustment suggestions for the monitoring parameters of the monitoring equipment according to the index values ​​of the output evaluation indicators and the priority of the evaluation indicators; and sending the adjustment suggestions to the operation and maintenance personnel so that the operation and maintenance personnel can adjust the monitoring equipment.

[0114] In one possible implementation, the adjustment unit 204 is used to: determine the adjustment information of the monitoring parameters of the monitoring device based on the priority of the evaluation indicator and the other related evaluation indicators when the evaluation indicator is related to other evaluation indicators; wherein the priority is positively correlated with the importance of the business processed by the monitoring device reflected by the evaluation indicator; and adjust the monitoring parameters of the monitoring device according to the determined adjustment information.

[0115] In one possible implementation, the adjustment unit 204 is used to: adjust the monitoring parameters of the monitoring equipment according to the adjustment information, so that the highest priority evaluation index reaches the optimal level, and other evaluation indicators other than the highest priority evaluation index meet the preset requirements.

[0116] In one possible implementation method, the evaluation indicators include at least one of the following: image quality evaluation, scene coverage evaluation, scene applicability evaluation, or target applicability evaluation.

[0117] In one possible implementation method, the image quality evaluation specifically includes at least one of the following: the side angle size, pitch angle size, rotation angle size, resolution size, light intensity, or target blur of the monitoring device.

[0118] In one possible implementation method, scene coverage evaluation specifically includes: confidence level of scene coverage of monitoring equipment from different angles.

[0119] In one possible implementation method, the scene suitability evaluation specifically includes at least one of the following: the screen occlusion of the monitoring equipment, background suitability, or site suitability.

[0120] In one possible implementation method, the target suitability evaluation specifically includes at least one of the following: target type ratio analysis, main traffic flow direction analysis, or capture location analysis of the monitoring equipment.

[0121] Based on the same technical concept, embodiments of this application provide an adjustment device 300 for a monitoring device, which may be, for example, a computing device. Figure 3 As shown, the adjustment device 300 of the monitoring equipment includes at least one processor 301 and a memory 302 connected to the at least one processor. In this embodiment, the specific connection medium between the processor 301 and the memory 302 is not limited. Figure 3 Taking the connection between processor 301 and memory 302 via a bus as an example, the bus can be divided into address bus, data bus, control bus, etc.

[0122] In this embodiment of the application, the memory 302 stores instructions that can be executed by at least one processor 301. By executing the instructions stored in the memory 302, at least one processor 301 can perform the above-mentioned method for adjusting the monitoring device.

[0123] The processor 301 serves as the control center of the adjustment device 300 in the monitoring equipment. It can connect to various parts of the computer equipment via various interfaces and lines, and performs resource settings by running or executing instructions stored in the memory 302 and calling data stored in the memory 302. Optionally, the processor 301 may include one or more determining units. The processor 301 may integrate an application processor and a modem processor. The application processor primarily handles the operating system, user interface, and applications, while the modem processor primarily handles wireless communication. It is understood that the modem processor may not be integrated into the processor 301. In some embodiments, the processor 301 and the memory 302 may be implemented on the same chip; in other embodiments, they may be implemented on separate chips.

[0124] Processor 301 can be a general-purpose processor, such as a central processing unit (CPU), digital signal processor, application-specific integrated circuit (ASIC), field-programmable gate array (FPGA), or other programmable logic device, discrete gate or transistor logic device, or discrete hardware component, capable of implementing or executing the methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this application can be directly manifested as being executed by a hardware processor, or executed by a combination of hardware and software modules within the processor.

[0125] Memory 302, as a non-volatile computer-readable storage medium, can be used to store non-volatile software programs, non-volatile computer-executable programs, and modules. Memory 302 may include at least one type of storage medium, such as flash memory, hard disk, multimedia card, card-type memory, random access memory (RAM), static random access memory (SRAM), programmable read-only memory (PROM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), magnetic storage, magnetic disk, optical disk, etc. Memory 302 can be any other medium capable of carrying or storing desired program code in the form of instructions or data structures that can be accessed by a computer, but is not limited thereto. In the embodiments of this application, memory 302 may also be a circuit or any other device capable of implementing storage functions for storing program instructions and / or data.

[0126] This application also provides a computer-readable storage medium storing a computer-executable program for causing a computer to perform the adjustment method of the monitoring device listed in any of the above methods.

[0127] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0128] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to this application. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1A device that provides the functions specified in one or more boxes.

[0129] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0130] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0131] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A method for adjusting a monitoring device, characterized in that, include: The system acquires multiple images collected by the monitoring device within a target time period. Each of the multiple images contains a target object, and the target objects contained in the multiple images correspond to the same target type. Each target type corresponds to at least one attribute feature. Identify various attribute features of the target object and the corresponding position of each attribute feature in the images from the multiple images; For each attribute feature, according to the time sequence of the acquisition time of the multiple images and the corresponding position of the attribute feature in the image, the attribute feature of the target object in the multiple images is integrated to form an attribute map corresponding to the attribute feature, wherein at least one attribute map corresponds one-to-one with at least one attribute feature; The at least one attribute graph is evaluated to obtain at least one evaluation index corresponding to the monitoring device; The monitoring equipment is adjusted according to at least one evaluation indicator.

2. The method according to claim 1, characterized in that, Adjusting the monitoring equipment according to the at least one evaluation index includes: Obtain the priority of the output evaluation indicators; the priority is related to the business to be performed by the monitoring device; Based on the output values ​​of the evaluation indicators and their priorities, the monitoring parameters of the monitoring equipment are adjusted.

3. The method according to claim 2, characterized in that, Adjusting the monitoring parameters of the monitoring equipment based on the output evaluation index values ​​and the priority of the evaluation index includes: Based on the output evaluation index values ​​and the priority of the evaluation index, adjustment suggestions for the monitoring parameters of the monitoring equipment are determined. The adjustment suggestions are sent to the operations and maintenance personnel so that they can adjust the monitoring equipment.

4. The method according to claim 2, characterized in that, The step of adjusting the monitoring parameters of the monitoring equipment based on the output evaluation index values ​​and the priority of the evaluation index includes: In the case of other evaluation indicators related to the evaluation indicator, the adjustment information of the monitoring parameters of the monitoring equipment is determined based on the evaluation indicator and the priority of the other related evaluation indicators; wherein, the priority is positively correlated with the importance of the business processed by the monitoring equipment as reflected by the evaluation indicator; The monitoring parameters of the monitoring equipment are adjusted based on the determined adjustment information.

5. The method according to claim 4, characterized in that, Adjusting the monitoring parameters of the monitoring device based on the determined adjustment information includes: Based on the adjustment information, the monitoring parameters of the monitoring equipment are adjusted so that the highest priority evaluation indicator reaches its optimal state, and other evaluation indicators besides the highest priority evaluation indicator meet the preset requirements.

6. The method according to any one of claims 1 to 5, characterized in that, The evaluation indicators include at least one of the following: Image quality evaluation, scene coverage evaluation, scene applicability evaluation, or target applicability evaluation.

7. The method according to claim 6, characterized in that, The image quality evaluation includes at least one of the following: the side angle size, pitch angle size, rotation angle size, resolution size, light intensity, or target blur of the monitoring device; The scene coverage evaluation includes: the confidence level of the scene coverage of the monitoring equipment from different angles; The scene suitability evaluation includes at least one of the following: the screen occlusion of the monitoring equipment, background suitability, or site suitability; The target suitability evaluation includes at least one of the following: target type ratio analysis, main traffic flow direction analysis, or capture location analysis of the monitoring equipment.

8. An adjustment device for a monitoring equipment, characterized in that, include: The acquisition unit is used to acquire multiple images collected by the monitoring device within a target time period, wherein each of the multiple images contains a target object, and the target objects contained in the multiple images correspond to the same target type; The determining unit is used to identify multiple attribute features of the target object and the corresponding position of each attribute feature in the image from the multiple images respectively; for each attribute feature, according to the time sequence of the acquisition time of the multiple images and the corresponding position of the attribute feature in the image, the attribute feature of the target object in the multiple images is integrated to form an attribute map corresponding to the attribute feature, wherein at least one attribute map corresponds one-to-one with at least one attribute feature; An evaluation unit is used to evaluate the at least one attribute graph to obtain at least one evaluation index corresponding to the monitoring device; An adjustment unit is used to adjust the monitoring equipment according to the at least one evaluation index.

9. A computer-readable storage medium, characterized in that, It includes computer-readable instructions that, when read and executed by a computer, implement the method as described in any one of claims 1 to 7.

Images