An integrated analysis method, system, storage medium and program product suitable for multi-scene dynamic images

Abstract

The application provides an integrated analysis method and system suitable for multi-scene dynamic images, a storage medium and a program product, and relates to the technical field of image equipment. The alien animal body is tracked. When tracking, the original volume information of the alien animal body in the actual space is calculated, and when the alien animal body cannot be tracked, all other alien animal bodies in the monitoring picture are acquired, and volume calculation strategies are respectively performed on them. Then, whether the difference between the sum of the other alien animal bodies and the original volume information is less than a difference threshold value is judged to screen out possible original alien animal bodies. If the difference meets the requirements, the other alien animal body with a volume sequence change trend of reduction is determined as the original alien animal body, and tracking is performed. The defects that the related tracking mode is easy to lose the target and misjudge are effectively made up, the target can be found back after the target is lost by relying on the volume-related judgment, the whole tracking process can be smoothly continued, and the tracking success rate of the illegal vehicle that randomly dumps the slag is increased.

Description

Technical Field

[0001] This application relates to the field of image equipment technology, and in particular to an integrated analysis method, system, storage medium and program product applicable to dynamic images in multiple scenes. Background Technology

[0002] Today, the illegal dumping of construction waste is a persistent problem that has severely impacted the city's ecological environment, appearance, and infrastructure.

[0003] Tracking vehicles illegally dumping construction waste presents numerous challenges. Firstly, the waste often partially obscures the vehicle's main structure, making previously easily identifiable features difficult to discern. Crucially, the dumping process causes significant deformation, such as body twisting and changes in the truck bed angle. This results in drastic changes in vehicle characteristics, making it difficult to match features even within a very short timeframe, significantly increasing the difficulty of monitoring. While attempting to reduce the sampling time interval for more accurate vehicle feature capture introduces new problems, it necessitates a significantly higher performance requirement for the entire monitoring system, both in terms of hardware and computing resources. This is prohibitively expensive for tracking illegal construction waste dumping.

[0004] In conclusion, the relevant technologies have a poor success rate in tracking vehicles illegally dumping construction waste. Summary of the Invention

[0005] This application provides an integrated analysis method, system, storage medium, and program product applicable to dynamic images in multiple scenarios, which can improve the success rate of tracking illegally dumped construction waste vehicles while keeping the increase in cost controllable.

[0006] Firstly, this application provides an integrated analysis method for dynamic images applicable to multiple scenes, comprising: acquiring monitoring images of different scenes through image acquisition devices deployed in multiple scenes; analyzing a preset area of ​​the acquired monitoring image to determine whether there are any foreign objects within the preset area, wherein a foreign object refers to an element within the preset area that has a preset difference threshold greater than or equal to that of other areas of the monitoring image; dividing the foreign object into several first sub-regions, and clustering the first sub-regions according to similarity and interval to obtain several second sub-regions; wherein the higher the similarity or the lower the interval, the higher the probability of clustering; and clustering according to a height ratio threshold. The alien object is divided into an upper region and a lower region. Second sub-regions within the upper region are added to the algorithm repository in descending order of shape complexity to determine if they possess pre-stored features related to construction waste. Similarly, second sub-regions within the lower region are added to the algorithm repository in ascending order of shape complexity to determine if they possess pre-stored features related to automotive parts. If any second sub-region within the upper region possesses pre-stored features related to construction waste, and any second sub-region within the lower region possesses pre-stored features related to automotive parts, the alien object is tracked. During tracking, other image acquisition devices in the corresponding scene are activated to capture images. Several target images containing alien objects are captured by other image acquisition devices positioned at different locations within the corresponding scene, each with a different shooting angle. A volume calculation strategy is initiated. The volume calculation strategy includes: calculating the original volume information of the alien object in actual space based on the height, angle, focal length of the other image acquisition devices, and the pixel size and position information of the alien object in the target image; periodically updating the volume data of the alien object at time intervals to form a volume sequence showing the change in alien object volume over time; acquiring all other alien objects within the monitored image when tracking is impossible; applying the volume calculation strategy to each of the other alien objects to obtain other volume information and other volume sequences; selecting two other volume information from all other volume information, defining them as the first other volume information and the second other volume information; determining whether the difference between the sum of the first other volume information and the second other volume information and the original volume information is less than a difference threshold; if it is less than the difference threshold, determining the change trend of the other volume sequences corresponding to the first other volume information and the second other volume information; identifying alien objects whose other volume sequences corresponding to the first other volume information or the second other volume information show a decreasing change trend as alien objects, tracking them, and modifying the original volume information accordingly.

[0007] By employing the aforementioned technical solution, image acquisition devices deployed in multiple scenes are used to acquire monitoring images of different scenarios and identify any unusual objects within them, thus identifying the objects of interest for subsequent processing. Next, these unusual objects are segmented into several first sub-regions, and then clustered according to similarity and interval to obtain several second sub-regions. Vehicles transporting construction waste are divided into several regions with distinct, singular, and large areas to facilitate the differentiation of vehicle and construction waste features, improving the overall accuracy and efficiency of the judgment. Considering that construction waste is generally at the top and has the most irregular shape, the second sub-regions within the upper region are placed into the algorithm repository in descending order of shape complexity to determine if they possess pre-stored construction waste features. Conversely, some car parts are located at the bottom and are often relatively regular and simple in shape; therefore, the second sub-regions within the lower region are placed into the algorithm repository in ascending order of shape complexity to determine if they possess pre-stored car part features. This approach fully utilizes the distribution characteristics and shape properties of construction waste and car parts for feature judgment, reducing computational difficulty and resource consumption. Then, the unusual objects are tracked. During tracking, other image acquisition devices in the corresponding scene are activated to capture several target images containing the alien object. These devices, distributed in different locations and with different shooting angles, capture multi-view images that can present the state of the alien object from all angles, providing a data foundation for subsequent volume calculation. Then, a volume calculation strategy is initiated to calculate the original volume information of the alien object in actual space. The volume data of the alien object is then periodically updated at time intervals, forming a volume sequence showing the change in the alien object's volume over time. This allows for real-time monitoring of the alien object's volume changes, providing crucial information for judging its behavioral state. When an alien object cannot be tracked, all other alien objects in the monitoring screen are acquired, and their volume calculation strategies are applied to each, yielding other volume information and other volume sequences. Then, two suitable other volume information are selected from all other volume information, designated as the first and second other volume information. By determining whether the difference between their sum and the original volume information is less than a difference threshold, the possible original alien object is filtered out. If the difference meets the requirements, further determine the change trend of other corresponding volume sequences, identify other alien animals whose volume sequence change trend is decreasing as the original alien animals, and track them. At the same time, use the corresponding volume information to modify the original volume information.In real-world scenarios involving the dumping of construction waste, this system effectively addresses numerous challenges, such as the obstruction of vehicles by the waste, the constantly changing shape of vehicles during dumping, and the complexity and difficulty in identifying vehicle characteristics. It overcomes the shortcomings of other tracking methods, which are prone to losing targets and making misjudgments. This system can accurately lock onto targets for continuous tracking, reduce computational resource consumption through reasonable area division and feature judgment, and even retrieve targets after they are lost by using volume-related judgments. This allows the entire tracking process to continue smoothly and increases the success rate of tracking illegally dumped construction waste vehicles.

[0008] In conjunction with some embodiments of the first aspect, in some embodiments, the difference threshold is positively correlated with the time elapsed since the loss of tracking of the alien body; the step of selecting two other volume information from all other volume information and determining them as the first other volume information and the second other volume information specifically includes: selecting two other volume information whose distance is less than a preset departure threshold from all other volume information and determining them as the first other volume information and the second other volume information; the preset departure threshold is positively correlated with the time elapsed since the loss of tracking of the alien body.

[0009] By adopting the above technical solution, considering the impact of dust interference and the passage of time—for example, the original volume of an object may change significantly due to changes in visibility after dust dissipates, and the uncertainty increases with time—correlating the difference threshold with the duration of time can better accommodate these uncertainties in volume changes caused by dust and time factors. This avoids misjudgment due to overly strict volume thresholds, thereby increasing the probability of identifying the original alien object from numerous other alien objects and creating favorable conditions for subsequent re-tracking. Simultaneously, two other volume information points with a distance less than a preset escape threshold are selected from all other volume information and designated as the first and second other volume information points. The preset escape threshold is positively correlated with the time elapsed since the loss of tracking of the alien object, taking into account the spatial distance factor that changes over time in dust interference scenarios. When dust causes the image to be blurry and object features to be difficult to distinguish clearly, vehicles may move away over time, increasing the distance between the debris and the vehicle. By positively correlating the preset escape threshold with the duration of time, the judgment standard for distance can be dynamically adjusted according to time. As time progresses, the distance restrictions are reasonably relaxed, allowing the selected alien objects corresponding to the two other volume information to better match the spatial changes of the original target in a dusty environment and after the vehicle may move. This avoids visual illusions caused by dust and distance changes caused by the vehicle moving away, which could lead to errors in distance judgment. As a result, possible alien objects can be screened more accurately, enhancing the accuracy of re-identifying the target.

[0010] In conjunction with some embodiments of the first aspect, in some embodiments, after the step of tracking the foreign object, the method further includes: generating a corresponding alarm strategy; determining whether the movement rhythm data of the foreign object conforms to a preset uniformity rule; if it conforms to the preset uniformity rule, determining whether the movement rhythm data is within the movement rhythm feature range in the illegal dumping of construction waste feature library; if it is within the movement rhythm feature range in the illegal dumping of construction waste feature library, determining it as illegal dumping of construction waste and activating the alarm strategy; if it is not within the movement rhythm feature range in the illegal dumping of construction waste feature library, returning to the step of tracking the foreign object until a preset time is reached or no foreign object is detected on the monitoring screen.

[0011] By adopting the above technical solution, when a vehicle is identified as a construction waste transport vehicle, an alarm strategy is pre-generated to improve the efficiency of subsequent alarms and further obtain its movement rhythm data. By judging whether the movement rhythm data conforms to a preset uniformity rule, cases that occasionally show similar characteristics but are not actually illegally dumping construction waste are filtered out. Even if it conforms to the uniformity rule, it is still necessary to further determine whether it is within the movement rhythm feature range of the illegal dumping behavior feature database. Only when all requirements are confirmed at each level is the alarm strategy activated. This makes the judgment of illegal dumping behavior more accurate, effectively avoiding misjudgments caused by some normal transportation behaviors being similar to illegal dumping characteristics, and reducing the manpower and time resources consumed by regulatory personnel to verify misjudgments.

[0012] In conjunction with some embodiments of the first aspect, in some embodiments, after determining whether the movement rhythm data conforms to a preset uniformity rule, the method further includes: if it does not conform to the preset uniformity rule, collecting all behavioral data of the abnormal animal from its appearance on the monitoring screen to the current moment, the behavioral data including: driving trajectory, stopping time, speed change, and turning frequency; clustering the behavioral data with preset normal behavioral pattern data; deleting the alarm strategy if the behavioral data is clustered by any normal behavioral pattern data; and returning to the step of tracking the abnormal animal to obtain the movement rhythm data of the abnormal animal if the behavioral data is not clustered by any normal behavioral pattern data, until a preset time is reached or there is no abnormal animal on the monitoring screen.

[0013] By employing the above technical solution, when the movement rhythm data is determined to be inconsistent with the preset uniformity rule, multi-dimensional behavioral data of the abnormal animal from its appearance to the present is collected. This data comprehensively reflects its behavioral characteristics. Next, the behavioral data is clustered with preset normal behavioral pattern data. Through this comparative clustering, it is possible to accurately determine whether the behavior is normal. If it is clustered with a normal pattern, it indicates a likely normal situation; deleting the alarm strategy can avoid false alarms, reduce unnecessary interference, and improve alarm accuracy. If it is not clustered, new data is continuously acquired, and the judgment is repeated until the conditions are met.

[0014] In conjunction with some embodiments of the first aspect, in some embodiments, if the behavioral data is not clustered by any normal behavioral pattern data, the method returns to the step of tracking the alien body and obtaining the movement rhythm data of the alien body until a preset time is reached or there is no alien body in the monitoring screen. The method further includes: if the behavioral data is not clustered by any normal behavioral pattern data after the preset time is reached or there is no alien body in the monitoring screen; outputting the behavioral data to a database to be determined; if the database to be determined is determined, adding a new normal behavioral pattern and determining the behavioral data as sub-data of the new normal behavioral pattern.

[0015] By adopting the above technical solution, when behavioral data has not been clustered and a preset time has elapsed or there are no foreign objects in the monitoring screen, it is output to the database to be determined. This data is retained for subsequent analysis and further confirmation. Once the database to be determined is finalized, new normal behavioral patterns are added and their data is treated as sub-data. This continuously expands the range of normal behavioral patterns, making the system's understanding of normal behavior more comprehensive, and thus enabling more accurate differentiation between normal and abnormal behavior in subsequent judgments.

[0016] In conjunction with some embodiments of the first aspect, in some embodiments, if the movement rhythm feature is not within the movement rhythm feature range in the illegal dumping of construction waste feature database, the step of tracking the foreign object is returned until a preset time is reached or no foreign object is detected on the monitoring screen; if the preset time is reached or no foreign object is detected on the monitoring screen, and the movement rhythm data is not within the movement rhythm feature range in the illegal dumping of construction waste feature database, the alarm strategy is deleted.

[0017] By adopting the above technical solution, if the movement rhythm data is not within the range of the illegal dumping behavior feature database, the alarm strategy is to continuously track until a preset time or the absence of foreign objects is met, after which the alarm is deleted. This is to take into account the possibility that special circumstances may cause the rhythm data to be temporarily abnormal but not due to illegal dumping behavior. The alarm deletion strategy can avoid false alarms caused by such situations, reduce invalid alarm information, and allow the monitoring system to focus on genuine illegal dumping behavior.

[0018] In conjunction with some embodiments of the first aspect, in some embodiments, the step of placing the second sub-regions within the upper region into the algorithm warehouse in descending order of shape complexity, and determining whether they have pre-stored slag features, specifically includes: selecting the corresponding algorithm from the algorithm warehouse according to the image acquisition device corresponding to the second sub-region; allocating computing resources according to the algorithm warehouse, and determining whether they have pre-stored slag features.

[0019] By adopting the above technical solution, after placing the second sub-regions according to their shape complexity from largest to smallest, and then selecting the appropriate algorithm based on the corresponding image acquisition device, the most suitable algorithm can be matched based on the device characteristics, improving the judgment efficiency. Then, computing resources are allocated according to the algorithm repository, ensuring that resources are allocated on demand and avoiding waste.

[0020] Secondly, this application provides an integrated analysis system for multi-scene dynamic images. The integrated analysis system for multi-scene dynamic images includes: one or more processors and a memory; the memory is coupled to one or more processors, and the memory is used to store computer program code, the computer program code including computer instructions, and one or more processors call the computer instructions to cause the integrated analysis system for multi-scene dynamic images to perform the method described in the first aspect and any possible implementation of the first aspect.

[0021] Thirdly, this application provides a computer program product containing instructions that, when run on an integrated analysis system for multi-scene dynamic images, causes the integrated analysis system for multi-scene dynamic images to perform the method described in the first aspect and any possible implementation thereof.

[0022] Fourthly, this application provides a computer-readable storage medium including instructions that, when executed on an integrated analysis system for multi-scene dynamic images, cause the integrated analysis system for multi-scene dynamic images to perform the method described in the first aspect and any possible implementation thereof.

[0023] One or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages:

[0024] 1. Using image acquisition devices deployed in multiple scenes, monitoring images of different scenes are acquired, and foreign objects are identified to pinpoint the objects of interest for subsequent processing. Next, the foreign objects are segmented into several first sub-regions, and these first sub-regions are clustered according to similarity and interval to obtain several second sub-regions. Vehicles transporting construction waste are divided into several regions with distinct, singular, and large areas to facilitate the differentiation of vehicle and construction waste features, improving the overall accuracy and efficiency of the judgment. Considering that construction waste is generally at the top and has the most irregular shape, the second sub-regions within the upper region are placed into the algorithm repository in descending order of shape complexity to determine if they possess pre-stored construction waste features. Conversely, some car parts are located at the bottom and are often relatively regular and simple in shape; therefore, the second sub-regions within the lower region are placed into the algorithm repository in ascending order of shape complexity to determine if they possess pre-stored car part features. This approach fully utilizes the distribution characteristics and shape properties of construction waste and car parts for feature judgment, reducing computational difficulty and resource consumption. Then, the foreign objects are tracked. During tracking, other image acquisition devices in the corresponding scene are activated to capture several target images containing the alien object. These devices, distributed in different locations and with different shooting angles, capture multi-view images that can present the state of the alien object from all angles, providing a data foundation for subsequent volume calculation. Then, a volume calculation strategy is initiated to calculate the original volume information of the alien object in actual space. The volume data of the alien object is then periodically updated at time intervals, forming a volume sequence showing the change in the alien object's volume over time. This allows for real-time monitoring of the alien object's volume changes, providing crucial information for judging its behavioral state. When an alien object cannot be tracked, all other alien objects in the monitoring screen are acquired, and their volume calculation strategies are applied to each, yielding other volume information and other volume sequences. Then, two suitable other volume information are selected from all other volume information, designated as the first and second other volume information. By determining whether the difference between their sum and the original volume information is less than a difference threshold, the possible original alien object is filtered out. If the difference meets the requirements, the changing trends of other corresponding volume sequences are further determined. Other alien objects whose volume sequence changing trend is decreasing are identified as the original alien objects and tracked. At the same time, the original volume information is modified using the corresponding volume information. In the actual scenario of dumping construction waste, facing many challenges such as construction waste obscuring vehicles, the vehicle's shape constantly changing with the dumping action, and the complexity and difficulty in identifying vehicle characteristics, this method effectively makes up for the shortcomings of related tracking methods, which are prone to losing targets and making misjudgments. It can accurately lock onto the target for continuous tracking, reduce computational resource consumption through reasonable area division and feature judgment, and even find the target after it is lost by judging volume-related factors, allowing the entire tracking process to continue smoothly and increasing the success rate of tracking illegal vehicles dumping construction waste.

[0025] 2. Considering the interference of dust and the impact of time, such as the significant changes in the original volume of an object due to visual alterations after dust dissipates, with increasing uncertainty over time, linking the difference threshold to the duration better accommodates these uncertainties caused by dust and time factors. This avoids misjudgments due to overly strict volume thresholds, thereby increasing the probability of identifying the original object from among numerous other objects and creating favorable conditions for subsequent re-tracking. Simultaneously, two other volume information points with a distance less than a preset escape threshold are selected from all other volume information and designated as the first and second other volume information points. The preset escape threshold is positively correlated with the time elapsed since the loss of tracking the object, taking into account the spatial distance factors changing over time in dust-affected scenarios. When dust blurs the image and makes object features difficult to distinguish, vehicles may move away over time, increasing the distance between the debris and the vehicle. By positively correlating the preset escape threshold with the duration, the distance judgment standard can be dynamically adjusted based on time. As time progresses, the distance restrictions are reasonably relaxed, allowing the selected alien objects corresponding to the two other volume information to better match the spatial changes of the original target in a dusty environment and after the vehicle may move. This avoids visual illusions caused by dust and distance changes caused by the vehicle moving away, which could lead to errors in distance judgment. As a result, possible alien objects can be screened more accurately, enhancing the accuracy of re-identifying the target.

[0026] 3. When a vehicle is identified as a construction waste transport vehicle, an alarm strategy is pre-generated to improve subsequent alarm efficiency and further obtain its movement rhythm data. By judging whether the movement rhythm data conforms to a preset uniformity rule, cases that occasionally show similar characteristics but are not actually illegally dumping construction waste are filtered out. Even if it conforms to the uniformity rule, it is still necessary to further determine whether it is within the movement rhythm feature range of the illegal dumping behavior feature database. Only when all the requirements are confirmed at each level is the alarm strategy activated. This makes the judgment of illegal dumping behavior more accurate, effectively avoiding misjudgments caused by some normal transportation behaviors being similar to illegal dumping characteristics, and reducing the manpower and time resources consumed by regulatory personnel to verify misjudgments. Attached Figure Description

[0027] Figure 1 This is a flowchart illustrating an integrated analysis method for dynamic images in multiple scenes, as described in this application.

[0028] Figure 2 This is another flowchart illustrating the integrated analysis method applicable to dynamic images in multiple scenes, as described in this application embodiment;

[0029] Figure 3This is another flowchart illustrating the integrated analysis method applicable to dynamic images in multiple scenes, as described in this application embodiment;

[0030] Figure 4 This is an exemplary hardware structure diagram of an integrated analysis system applicable to dynamic images in multiple scenes, as described in this application. Detailed Implementation

[0031] The terminology used in the following embodiments of this application is for the purpose of describing particular embodiments only and is not intended to be limiting of this application. As used in the specification and appended claims of this application, the singular expressions “a,” “an,” “the,” “the,” “the,” and “this” are intended to include the plural expressions as well, unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used in this application refers to and includes any or all possible combinations of one or more of the listed items.

[0032] Hereinafter, the terms "first" and "second" are used for descriptive purposes only and should not be construed as implying or suggesting relative importance or implicitly indicating the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature, and in the description of the embodiments of this application, unless otherwise stated, "multiple" means two or more.

[0033] Please see Figure 1 , Figure 1 This is a flowchart illustrating an integrated analysis method for dynamic images in multiple scenes, as described in this application.

[0034] S101. Acquire monitoring images of different scenes through image acquisition devices deployed in multiple scenes;

[0035] Among them, multiple scenarios represent multiple scenarios of different types and environments, covering various areas that may need to be monitored;

[0036] It should be noted that this application has the capability to perform the following steps in multiple scenarios. However, for ease of understanding, the explanations of subsequent steps S102 to S112 are all based on the situation occurring in one scenario.

[0037] In some preferred embodiments, for a given scene, only one image acquisition device needs to be activated, and this activated device should be the one with the largest shooting area. This is because the key task of steps S101 and subsequent steps S102 up to the current step is to determine whether any foreign objects appear in the scene. By selecting the image acquisition device with the largest shooting area, the scene coverage can be maximized, capturing image information more comprehensively and efficiently. This provides strong data support for accurately determining the presence of foreign objects, avoiding the omission of potential targets due to limited shooting area, and ensuring the successful completion of this stage of the task.

[0038] S102. Analyze the preset area of ​​the acquired monitoring screen to determine whether there are any foreign objects in the preset area. Foreign objects refer to objects in the preset area that have a preset difference threshold or higher than the elements in other areas of the monitoring screen.

[0039] The preset area refers to a specific area pre-defined in the monitoring screen. This area is the focus of the analysis; for example, it could be a specific rectangular area in the screen. An abnormal object refers to an element within the preset area that differs from other areas of the monitoring screen by a preset difference threshold. This difference threshold is a pre-defined standard used to determine whether an object is abnormal; for example, objects with differences in color, brightness, shape, etc., reaching a certain level can be considered abnormal objects.

[0040] In some embodiments, features are then extracted and analyzed from the elements within the area. By comparing the features with those of other areas of the monitoring screen, the degree of difference is calculated. When the difference exceeds a preset difference threshold, it is determined that there is an alien object in the area, providing a basis for further processing of the alien object.

[0041] In some specific embodiments, the monitoring image is first preprocessed using image processing algorithms, such as noise reduction and contrast enhancement; then features of a preset area and other areas are extracted, such as color histograms and texture features; next, the difference between the two is calculated and compared with a preset difference threshold. If the difference exceeds the threshold, it is determined that there is an abnormal animal.

[0042] It should be noted that when determining whether there are any foreign objects within a preset area, both horizontal and vertical comparison methods can be used. Horizontal comparison involves comparing the preset area with other areas within the same monitoring frame to check for differences. For example, when animals or people appear in the frame, the preset area containing these elements often shows significant differences in color, shape, and movement compared to other regular areas in the frame, thus indicating the presence of foreign objects. Vertical comparison, on the other hand, involves observing adjacent monitoring frames to see if there are any differences in features between the preset area and the same area at different times. For instance, if a previously flat area suddenly shows newly added mounds of soil in subsequent adjacent monitoring frames, it means that the area has changed. This comparison with the same area at different times can also determine whether there are foreign objects within the preset area.

[0043] S103. Divide the alien body into several first sub-regions, and cluster the first sub-regions according to similarity and interval to obtain several second sub-regions; the higher the similarity or the lower the interval, the higher the probability of clustering.

[0044] The first sub-region is a smaller regional unit obtained by segmenting the alien body. The purpose of segmentation is to analyze the characteristics of the alien body in more detail. The second sub-region is the region obtained by clustering the first sub-region according to similarity and interval. Similarity is used to measure the degree of similarity between two first sub-regions, and interval represents the spatial distance between two first sub-regions. The higher the similarity or the lower the interval, the higher the probability of clustering.

[0045] In some embodiments, to better understand the structure and characteristics of the exotic animal, it is segmented into multiple first sub-regions. Then, by calculating the similarity and interval between each first sub-region, the first sub-regions with high similarity and small intervals are clustered into second sub-regions. This allows the characteristics of the exotic animal to be classified and integrated, facilitating more accurate judgment and identification in subsequent processes.

[0046] In some specific embodiments, the alien body is divided into multiple first sub-regions based on a grid; then the feature vector of each first sub-region is calculated, such as color features, shape features, etc.; then the similarity and interval are calculated based on the feature vectors, and a clustering algorithm is used to cluster the first sub-regions with high similarity and low interval into second sub-regions, which is not limited here.

[0047] This allows the vehicles transporting construction waste to be divided into several distinct, singular, and large areas.

[0048] It should be noted that in some embodiments, the number of first sub-regions is considered and selected. This is because the number of first sub-regions varies among different types of organisms. Some organisms, due to their inherent characteristics, have relatively fewer first sub-regions, which often indicates a smaller overall volume for that organism. In practice, vehicles that do not meet the criteria for transporting construction waste are deemed to have relatively limited value and will not undergo further processing.

[0049] S104. Divide the alien body into an upper region and a lower region according to the height ratio threshold. Put the second sub-region inside the upper region into the algorithm warehouse in order of shape complexity from large to small, and determine whether it has the pre-stored slag features. Put the second sub-region inside the lower region into the algorithm warehouse in order of shape complexity from small to large, and determine whether it has the pre-stored car part features.

[0050] In some embodiments, shape complexity is the ratio of the perimeter of an object's boundary to its area. An irregularly shaped pile of slag with many bumps and depressions will have a very convoluted boundary, resulting in a much larger perimeter compared to a simpler shape of the same area. For the same area, objects with complex shapes have longer perimeters. Therefore, by calculating the ratio of perimeter to area, a larger ratio indicates a more convoluted boundary, a more irregular shape, and thus greater complexity.

[0051] Then, for the second sub-region of the upper region, they are sorted in descending order of shape complexity. A feature extraction algorithm is used to combine these features into a feature vector. This feature vector is then input into a support vector machine classifier in an algorithm repository, which has been trained using a large amount of construction waste sample data. The classifier calculates the similarity between the feature vector and pre-stored construction waste features; if the similarity is greater than a set threshold, the second sub-region is determined to have construction waste features. For the second sub-region of the lower region, they are also sorted in ascending order of shape complexity, and feature vectors are extracted and input into another decision tree-based classifier in the algorithm repository. This classifier is trained using automotive parts sample data, and judgment rules determine whether it has automotive parts features.

[0052] As can be seen, considering that construction waste is generally located at the top and has the most irregular shape, the second sub-region within the upper region is placed into the algorithm repository in descending order of shape complexity to determine whether it possesses pre-stored construction waste features. Conversely, some car parts are located at the bottom and tend to have relatively regular and simple shapes. Therefore, the second sub-region within the lower region is placed into the algorithm repository in ascending order of shape complexity to determine whether it possesses pre-stored car part features. This approach fully leverages the distribution characteristics and shape properties of both construction waste and car parts for feature determination, reducing computational complexity and resource consumption.

[0053] In some embodiments, step S104 specifically includes:

[0054] S1041. Based on the image acquisition device corresponding to the second sub-region, select the corresponding algorithm from the algorithm repository;

[0055] The algorithm repository is a database storing various algorithms for image analysis and feature recognition, including algorithms designed for different scenarios, features, and image acquisition devices. Selecting the appropriate algorithm means choosing the most suitable algorithm from the repository based on the characteristics of the image acquisition device corresponding to the second sub-region, in order to analyze and process the image and identify specific features (such as slag features).

[0056] In some embodiments, the monitoring system may deploy multiple different types of image acquisition devices, such as ordinary visible light cameras and thermal imaging cameras, which capture images of the object from different angles and in different ways. Once the second sub-region is obtained, the first step is to determine which image acquisition device acquired the image of that sub-region. Then, based on the parameters and characteristics of that image acquisition device, a matching algorithm is searched from the algorithm repository.

[0057] S1042. Based on the algorithm repository, allocate computing resources and determine whether the soil features are pre-stored.

[0058] In some embodiments, after a suitable algorithm is selected, resources such as CPU time, GPU cores, and memory space need to be allocated from the system's computing resource pool according to the algorithm's complexity and computational requirements. After allocating computing resources, the image of the second sub-region is input into the algorithm for processing. The algorithm extracts features from the image and compares them with pre-stored features of construction waste. The similarity between features is calculated or a classification algorithm is used for judgment. If the similarity exceeds a certain threshold or the feature is classified as construction waste, the second sub-region is determined to have pre-stored features of construction waste; otherwise, it is determined not to have them.

[0059] As can be seen, after placing the second sub-regions according to their shape complexity from largest to smallest, the corresponding image acquisition device is used to select the most suitable algorithm. This approach matches the most suitable algorithm based on the device's characteristics, improving judgment efficiency. Then, computing resources are allocated according to the algorithm repository, ensuring that resources are allocated on demand and avoiding waste.

[0060] S105. When any second sub-region within the upper region has pre-stored slag features and any second sub-region within the lower region has pre-stored automobile part features, track the foreign animal body.

[0061] This step is performed after determining that the second sub-regions of the upper and lower regions have pre-stored characteristics of construction waste and vehicle parts, respectively. The scenario is that the alien object is likely a vehicle transporting construction waste, and continuous monitoring of it is required.

[0062] In some specific embodiments, target detection algorithms are used to determine the position of the alien object in the monitoring screen; then, tracking algorithms such as Kalman filtering are used to predict its next position based on its historical position and movement state; finally, the position information of the alien object is continuously updated in subsequent monitoring screens to achieve continuous tracking. It is understood that other methods can also be used to implement this step, such as using feature-matching-based tracking algorithms, etc., which are not limited here.

[0063] S106. During tracking, other image acquisition devices in the corresponding scene are activated to acquire several target images containing the alien object. The other image acquisition devices are arranged in different positions in the corresponding scene with different shooting angles, and a volume calculation strategy is initiated. The volume calculation strategy includes: calculating the original volume information of the alien object in the actual space based on the height, angle, focal length of the other image acquisition devices and the pixel size and position information of the alien object in the target image; periodically updating the volume data of the alien object at time intervals to form a volume sequence of the alien object's volume changing over time.

[0064] In this context, "other image acquisition devices" refers to image acquisition devices deployed at different locations within the corresponding scene, with different shooting angles, in addition to the initial device acquiring the monitoring image. These devices are used to acquire information about the alien object from multiple perspectives. The target image refers to the image or video clip containing the alien object captured by the other image acquisition devices. The volume calculation strategy is a set of methods and procedures used to calculate the volume of the alien object in actual space. Height, angle, and focal length are parameters of the other image acquisition devices. Pixel size and position information refer to the pixel size and position of the alien object in the target image. The original volume information is the calculated initial volume value of the alien object in actual space. The volume sequence refers to a sequence formed by periodically updated alien object volume data, used to reflect changes in volume over time.

[0065] In some specific embodiments, a three-dimensional coordinate system is first established based on the installation location and parameters of other image acquisition devices. Then, using the principle of stereo vision, the position and size of the alien object in the three-dimensional coordinate system are calculated using the pixel information of the alien object in multiple target images. Next, based on the calculated size information, the original volume information is calculated using volume calculation formulas (such as using corresponding geometric formulas for regular objects and voxelization methods for irregular objects). Finally, the above steps are repeated at set time intervals to update the volume data, forming a volume sequence. Optionally, structured light three-dimensional scanning technology is first used to acquire three-dimensional point cloud data of the alien object by emitting and receiving light through other image acquisition devices. Then, the three-dimensional point cloud data is processed and analyzed to calculate the volume of the alien object. Then, time series analysis methods are used to periodically update the volume data at time intervals to form a volume sequence, which is not limited here.

[0066] S107. In the event that the alien animal cannot be tracked, obtain all other alien animals in the monitoring screen;

[0067] It should be noted that the core principle of the tracking algorithm lies in first performing target recognition. In practical applications, if a situation arises where it is impossible to track an object, it can be determined that target recognition cannot be achieved. This phenomenon is closely related to the special circumstances of vehicles transporting construction waste during the dumping operation. When a vehicle is dumping, it undergoes extremely significant deformation, specifically manifested as the twisting of the vehicle body structure and a large change in the angle of the cargo box. These changes cause drastic alterations to the vehicle's originally stable and easily identifiable features, resulting in significant differences in the features exhibited by the vehicle at different times. This makes it difficult for the tracking algorithm to continuously and accurately identify the target under such drastic feature changes. Therefore, in step S105, the tracking operation cannot continue.

[0068] It's important to understand that using the vehicle for dumping construction waste is illegal, and drivers will remove their license plates before dumping. This means that in practice, tracking is primarily based on the vehicle's own characteristics.

[0069] The principle and process of identifying foreign animals in this step are similar to those in step S102. The relevant principles and processes can be referred to in step S102, and are not limited here.

[0070] It should be noted that in practical applications, when the aforementioned situation of being unable to continuously track occurs, at least two new objects will appear on the monitoring screen. The first is the vehicle. Although it is essentially the same vehicle that was previously being tracked, the vehicle itself undergoes significant changes during the dumping operation, such as body structural distortion and changes in the angle of the cargo bed, making its characteristics completely different from before. Therefore, it will be considered a different situation during tracking. Moreover, the differences in vehicle characteristics are particularly prominent during the dumping process, which will not be elaborated on here. Even after the dumping process is completed, although the differences in vehicle characteristics decrease relatively, the significant changes in characteristics throughout the entire tracking process greatly affect the continuity of tracking. The second is the added pile of construction waste.

[0071] S108. Perform volume calculation strategies on other different animal bodies to obtain other volume information and other volume sequences;

[0072] It should be noted that the principle and process of this step are similar to those of step S106. You can refer to step S106 here, and it will not be repeated here.

[0073] S109. Select two other volume information from all other volume information and determine them as the first other volume information and the second other volume information;

[0074] Among them, the first other volume information and the second other volume information are two arbitrary volume values ​​selected from all other volume information, which are used for subsequent comparison and judgment.

[0075] In some embodiments, two volume information items are selected from the volume information of all other alien objects as the first other volume information and the second other volume information. The selection of these two volume information items is the basis for subsequent determination of whether other alien objects are related to the original tracked alien object. By comparing and analyzing their volume information with that of the original tracked alien object, possible targets can be found.

[0076] S110. Determine whether the difference between the sum of the first other volume information and the second other volume information and the original volume information is less than the difference threshold.

[0077] This step is performed to further determine whether these other alien objects might be previously lost tracking objects. In some embodiments, the first other volume information and the second other volume information are summed, and then the difference between this sum and the original volume information is calculated. This difference is then compared to a pre-set difference threshold. If the difference is less than the threshold, it means that from a volume perspective, these two other alien objects may be related to the originally tracked alien object, providing a basis for further confirmation; if the difference is greater than or equal to the threshold, it can be basically determined that they are not closely related, and these two other alien objects can be excluded.

[0078] S111. If the difference is less than the threshold, then determine the changing trend of the other volume sequences corresponding to the first other volume information and the second other volume information.

[0079] It should be noted that in actual use, vehicles involved in illegal dumping of construction waste are usually characterized by an exposed top, meaning they lack a tarpaulin or are semi-enclosed. This phenomenon stems from several factors specific to the actual scenario. In the specific context of illegal dumping, speed is often the primary consideration for violators. Disassembling and reassembling a tarpaulin typically takes 2-3 hours, which is clearly an unacceptable time cost for violators eager to complete the illegal dumping quickly. Furthermore, in conventional vehicles with concealed tops, valuables may be stored inside during unloading, leading to issues like moisture damage or loss. However, for vehicles illegally dumping construction waste, their sole purpose is to quickly dump the waste; they are unconcerned about whether items inside will be damaged or damp due to the lack of a tarpaulin or other protection, nor are they worried about loss. Therefore, vehicles without tarpaulins or semi-enclosed vehicles with exposed tops become their preferred option for carrying out illegal activities, satisfying their need for quick and easy illegal dumping.

[0080] Therefore, in actual use, the situation where the sum of the volume of the vehicle and the dumped material in a fully enclosed vehicle exceeds the original volume information will not be considered.

[0081] When the difference between the sum of the first and second other volume information and the original volume information is less than a difference threshold, it is considered valid. This is because in actual scenarios involving the transportation and dumping of construction waste, the initially tracked alien objects (such as vehicles transporting construction waste) undergo changes (e.g., the vehicle begins dumping construction waste, which gradually accumulates into new shapes). These changed shapes manifest as other alien objects (one could be the vehicle itself during dumping, and the other could be the dumped construction waste). Their volume information should have a logical connection to the original alien object (the transport vehicle and the entire load of construction waste). From a volume perspective, a difference within a reasonable range (less than the difference threshold) is only a preliminary condition. Further analysis of the volume changes over time is needed to more deeply verify whether these other alien objects are derivatives of the original alien object produced under specific actions (such as dumping construction waste). For example, during the normal dumping of construction waste, the volume of the vehicle itself may show a specific trend of change due to factors such as changes in the angle of the truck bed (such as appearing to gradually decrease in volume from certain perspectives), while the volume of the dumped construction waste will show a gradual increasing trend as dumping continues, proving that they are closely related to the original foreign object and are manifestations of it at different stages and in different forms.

[0082] S112. Other animal bodies whose change trend of other volume sequence corresponding to the first other volume information or the second other volume information is decreasing are identified as other animal bodies and tracked. The corresponding volume information is used to modify the original volume information.

[0083] In this context, "alien object" refers to the target object that was previously tracked but later lost track of. Identifying it as an alien object means determining that the alien object corresponding to this other volume sequence is the same as the previously lost alien object. Tracking involves continuing to monitor and record its position, movement status, and other information. The original volume information is the volume value of the initially tracked alien object, calculated earlier. The corresponding volume information refers to the latest volume information of this other alien object currently identified as the original alien object. Changing the original volume information involves updating the previous original volume information with the latest volume information to ensure the accurate reflection of the alien object's volume data.

[0084] The scenario involves ultimately confirming and resuming tracking of previously lost alien objects. In some embodiments, when other volume sequences corresponding to a particular alien object show a decreasing trend, this alien object is determined to be the previously lost object by combining the previously established criteria such as volume difference. The tracking mechanism is then restarted to acquire its position, movement status, and other information in real time. Simultaneously, the current volume information of this confirmed alien object is used to update the previously recorded original volume information. Since its volume may have changed during the period of lost tracking, the updated data allows subsequent analysis (such as volume change trend analysis) to be based on more accurate data, ensuring the continuity and accuracy of the entire monitoring process.

[0085] In some preferred embodiments, considering the need for complete vehicle information and comprehensive behavioral trajectory, a strategy is adopted to overlay features from the current scenario (obtained in step S105) with features from other scenarios. This combination generates a complete sequence of vehicle information, providing detailed representation of vehicle behavior patterns and facilitating quick and accurate vehicle location. For example, in a real-world scenario, a vehicle might initially drive normally in one scenario, with its license plate still mounted on the vehicle body, without removing it. However, when it moves to another scenario, it might illegally remove its license plate and dump construction waste. In such cases, only by associating and integrating the information related to the vehicle from various scenarios can complete information about the vehicle be formed, fully reconstructing its entire behavioral process.

[0086] In addition, in some other preferred embodiments, in order to obtain more detailed vehicle-related information, monitoring information from other platforms is obtained, features are extracted using the same method, and overlaid according to different scenarios to obtain a more complete sequence.

[0087] As can be seen, by utilizing image acquisition devices deployed in multiple scenes to obtain monitoring images of different scenes and identifying foreign objects within them, the focus of subsequent processing is narrowed down. Next, the foreign objects are segmented into several first sub-regions, and these first sub-regions are clustered according to similarity and interval to obtain several second sub-regions. Vehicles transporting construction waste are divided into several regions with distinct, singular, and large areas to facilitate the differentiation of vehicle and construction waste features, improving the overall accuracy and efficiency of the judgment. Considering that construction waste is generally at the top and has the most irregular shape, the second sub-regions within the upper region are placed into the algorithm repository in descending order of shape complexity to determine if they possess pre-stored construction waste features. Conversely, some car parts are located at the bottom and are often relatively regular and simple in shape; therefore, the second sub-regions within the lower region are placed into the algorithm repository in ascending order of shape complexity to determine if they possess pre-stored car part features. This approach fully utilizes the distribution characteristics and shape properties of the construction waste and car parts for feature judgment, reducing computational difficulty and resource consumption. Then, the foreign objects are tracked. During tracking, other image acquisition devices in the corresponding scene are activated to capture several target images containing the alien object. These devices, distributed in different locations and with different shooting angles, capture multi-view images that can present the state of the alien object from all angles, providing a data foundation for subsequent volume calculation. Then, a volume calculation strategy is initiated to calculate the original volume information of the alien object in actual space. The volume data of the alien object is then periodically updated at time intervals, forming a volume sequence showing the change in the alien object's volume over time. This allows for real-time monitoring of the alien object's volume changes, providing crucial information for judging its behavioral state. When an alien object cannot be tracked, all other alien objects in the monitoring screen are acquired, and their volume calculation strategies are applied to each, yielding other volume information and other volume sequences. Then, two suitable other volume information are selected from all other volume information, designated as the first and second other volume information. By determining whether the difference between their sum and the original volume information is less than a difference threshold, the possible original alien object is filtered out. If the difference meets the requirements, the changing trends of other corresponding volume sequences are further determined. Other alien objects whose volume sequence changing trend is decreasing are identified as the original alien objects and tracked. At the same time, the original volume information is modified using the corresponding volume information. In the actual scenario of dumping construction waste, facing many challenges such as construction waste obscuring vehicles, the vehicle's shape constantly changing with the dumping action, and the complexity and difficulty in identifying vehicle characteristics, this method effectively makes up for the shortcomings of related tracking methods, which are prone to losing targets and making misjudgments. It can accurately lock onto the target for continuous tracking, reduce computational resource consumption through reasonable area division and feature judgment, and even find the target after it is lost by judging volume-related factors, allowing the entire tracking process to continue smoothly and increasing the success rate of tracking illegal vehicles dumping construction waste.

[0088] In actual use, the dumping of construction waste generates a large amount of dust. The dust permeates the area around the vehicles and the construction waste, making the boundary between the vehicles and the construction waste indistinct in the target image. Although they are actually distinguishable, they are difficult to discern from a visual perspective, interfering with the identification of other objects. At the same time, the dust causes the volume of other objects to change in ways that should not occur, making it even more difficult to re-identify them.

[0089] In this case, a fusion operation (allowing time extension) needs to be performed on the time period covered by steps S107 to S112. With this fusion change in time, the related requirements closely associated with it naturally need to be adjusted accordingly. Therefore, in some preferred embodiments, step S109 is replaced by selecting two other volume information from all the other volume information that are less than a preset escape threshold, and determining them as the first other volume information and the second other volume information; the preset escape threshold is positively correlated with the time elapsed from the loss of tracking of the alien object to the present.

[0090] This step is performed when the monitoring system loses tracking of an alien object and needs to search for a potentially related target from other objects in the monitoring screen. In some embodiments, after losing tracking of an alien object, multiple other objects may exist in the screen, each with corresponding volume information. To find clues that may be related to the original alien object from this information, two suitable pieces of information need to be selected from all the other volume information. Here, the distance between the other volume information is compared, and the two pieces of information with a distance less than a preset escape threshold are identified as the first and second other volume information. Moreover, the preset escape threshold is positively correlated with the time elapsed since the loss of tracking of the alien object. This is because over time, the alien object and its surrounding environment may undergo more changes, and the object's volume may fluctuate more significantly. For example, in a short period of time, the difference in volume between two objects may be small, and the preset escape threshold can be set relatively small; however, after a longer period of time, the state of the object may change significantly, and the allowable range of volume difference should be expanded accordingly. Therefore, the preset escape threshold should increase with the increase in time.

[0091] In step S110, the difference threshold is positively correlated with the time elapsed from the loss of tracking of the alien body to the present moment;

[0092] In some embodiments, when the monitoring system loses tracking of the alien object, multiple other objects will appear on the screen. It is necessary to determine whether these other objects are likely the original alien object or closely related to it by comparing the difference between their volume information and the original volume information of the alien object. The difference threshold is positively correlated with the time elapsed since the loss of tracking. This is because the alien object may undergo more changes over time; for example, some debris may dissipate into the air. Therefore, in a short period, the difference between the volume of other alien objects and the original alien object may be small, and the difference threshold can be set relatively small. However, after a longer period, the allowable range of volume differences should be expanded accordingly to accommodate greater changes that the alien object may undergo. Therefore, the difference threshold should increase with the increase in time.

[0093] It is evident that considering the interference of dust and the effects of time, such as the significant changes in the original volume of an object due to changes in visibility after dust dissipates, with the uncertainty increasing over time, linking the difference threshold to the duration better accommodates these uncertainties in volume changes caused by dust and time factors. This avoids misjudgments due to overly strict volume thresholds, thereby increasing the probability of identifying the original alien object from numerous other alien objects and creating favorable conditions for subsequent re-tracking. Simultaneously, two other volume information points with distances less than a preset escape threshold are selected from all other volume information and designated as the first and second other volume information points. The preset escape threshold is positively correlated with the duration elapsed since the loss of tracking the alien object, taking into account the spatial distance factors that change over time in dust-affected scenarios. When dust causes blurred images and makes object features difficult to distinguish clearly, vehicles may move away over time, increasing the distance between the debris and the vehicle. By positively correlating the preset escape threshold with the duration, the distance judgment standard can be dynamically adjusted based on time. As time progresses, the distance restrictions are reasonably relaxed, allowing the selected alien objects corresponding to the two other volume information to better match the spatial changes of the original target in a dusty environment and after the vehicle may move. This avoids visual illusions caused by dust and distance changes caused by the vehicle moving away, which could lead to errors in distance judgment. As a result, possible alien objects can be screened more accurately, enhancing the accuracy of re-identifying the target.

[0094] In the above embodiment, how to track vehicles transporting construction waste is achieved. However, in actual use, it is necessary to find out whether these vehicles are dumping construction waste.

[0095] Please see Figure 2 , Figure 2 This is another flowchart illustrating the integrated analysis method applicable to dynamic images in multiple scenes, as described in this application embodiment;

[0096] Therefore, in some preferred embodiments, step S104 is followed by:

[0097] S201. Generate the corresponding alarm policy;

[0098] Among them, the alarm strategy refers to a series of response rules and measures formulated for specific abnormal situations, which are used to issue alarms in a timely manner and take corresponding actions when an anomaly is detected.

[0099] In some embodiments, if a vehicle transporting construction waste is detected, an alarm strategy is generated accordingly, taking into account information such as the current time and the specific location.

[0100] S202. Determine whether the movement rhythm data of the alien animal conforms to the preset uniformity rule;

[0101] Among them, the preset uniformity rule is a pre-defined standard used to measure whether the movement rhythm of an animal conforms to a certain uniform and stable pattern. For example, it stipulates that during normal driving, the speed fluctuation range of a vehicle should not exceed a certain percentage, and the pause time should not be too long.

[0102] In some embodiments, the monitoring system collects real-time movement data of foreign objects (vehicles transporting construction waste), such as vehicle position and speed. Then, it calculates the vehicle's movement rhythm data based on this data. This movement rhythm data is compared with a preset uniformity rule to determine whether the vehicle's movement conforms to a normal, uniform pattern. If the vehicle's movement rhythm conforms to the preset uniformity rule, it indicates that the vehicle's driving state is relatively stable, the data has a certain degree of reliability, and can be processed subsequently; if it does not conform, there may be an anomaly, requiring further analysis.

[0103] In some specific embodiments, the vehicle's speed and acceleration are calculated based on changes in location information. The speed and acceleration data over a period of time are then processed to form a movement rhythm data sequence. Next, a range for speed fluctuations and a threshold for acceleration are set according to a preset uniformity rule. If all data falls within the rule's range, it is determined to conform to the preset uniformity rule; otherwise, it is determined not to conform.

[0104] If the preset uniformity rule is met, proceed to step S203; otherwise, proceed to step S206.

[0105] S203. If the preset uniformity rule is met, determine whether the movement rhythm data is within the movement rhythm feature range in the feature library of illegal dumping of construction waste.

[0106] The "Illegal Dumping of Construction Waste" feature database is a large database storing characteristic data related to illegal dumping of construction waste. It includes various behavioral characteristics of vehicles involved in illegal dumping, such as movement rhythm, dwell time, and location information. The "movement rhythm feature range" refers to a range of feature values ​​defined in the database for vehicle movement rhythm, representing the possible variations in vehicle movement rhythm during illegal dumping.

[0107] After initially determining that the vehicle's movement is relatively normal, further analysis is conducted to determine if there is a possibility of illegal dumping of construction waste. In some embodiments, when the monitoring system determines that the movement rhythm data of the abnormal object conforms to a preset uniformity rule, it compares this movement rhythm data with the movement rhythm feature intervals in the illegal dumping behavior feature database. If the movement rhythm data falls within the feature interval, it indicates that the vehicle's movement rhythm has some typical characteristics of illegal dumping behavior, and there is suspicion of illegal dumping; if it does not fall within the interval, it indicates that the vehicle's movement rhythm does not match the typical characteristics of illegal dumping behavior, and the possibility of illegal dumping is temporarily ruled out.

[0108] In some specific embodiments, relevant data on movement rhythm characteristic intervals are extracted from the feature database of illegal dumping of construction waste, including minimum and maximum speeds, minimum and maximum pause times, etc. The movement rhythm data of the illegally dumped objects collected in real time are processed, and corresponding key indicators such as speed and pause times are extracted. Then, these key indicators are compared with the boundary values ​​of the characteristic intervals. If all indicators are within the characteristic intervals, the movement rhythm data is determined to be within the movement rhythm characteristic interval; otherwise, it is determined not to be within it.

[0109] If the movement rhythm characteristic range is within the feature range of the illegal dumping of construction waste in the feature database, proceed to step S204; otherwise, proceed to step S205.

[0110] S204. If the movement rhythm feature range is within the feature database of illegal dumping of construction waste, it is determined to be illegal dumping of construction waste, and the alarm strategy is activated.

[0111] S205. If the movement rhythm feature is not within the range of the feature database of illegal dumping of construction waste, return to the step of tracking the abnormal body until the preset time is reached or there is no abnormal body on the monitoring screen.

[0112] As can be seen, when a vehicle is identified as a construction waste transport vehicle, an alarm strategy is pre-generated to improve the efficiency of subsequent alarms and further obtain its movement rhythm data. By judging whether the movement rhythm data conforms to a preset uniformity rule, cases that occasionally show similar characteristics but are not actually illegally dumping construction waste are filtered out. Even if it conforms to the uniformity rule, it is still necessary to further determine whether it is within the movement rhythm feature range of the illegal dumping behavior feature database. Only when all the requirements are confirmed at each level is the alarm strategy activated. This makes the judgment of illegal dumping behavior more accurate, effectively avoiding misjudgments caused by some normal transportation behaviors being similar to illegal dumping characteristics, and reducing the manpower and time resources consumed by regulatory personnel to verify misjudgments.

[0113] S206. If the preset uniformity rule is not met, collect all behavioral data of the alien body from the time it appears on the monitoring screen to the current time. The behavioral data includes: driving trajectory, stopping time, speed change, and turning frequency.

[0114] S207. Cluster the behavioral data with preset normal behavioral pattern data;

[0115] In some embodiments, after obtaining behavioral data of an abnormal animal, the monitoring system performs cluster analysis on it against preset normal behavior pattern data. The system calculates the similarity between the behavioral data and each normal behavior pattern data, and classifies the behavioral data into the closest normal behavior pattern category based on the similarity level. If the behavioral data can be clustered by a certain normal behavior pattern data, it indicates that the vehicle's behavior as a whole conforms to a normal transportation pattern.

[0116] In some specific embodiments, preset normal behavior pattern data are used as cluster centers, and the collected behavior data are used as samples to be classified. The distance (e.g., Euclidean distance) between each behavior data sample and each cluster center is calculated. Then, the behavior data sample is assigned to the normal behavior pattern category represented by the nearest cluster center according to the distance, which is not limited here.

[0117] S208. If the behavioral data is clustered by any normal behavioral pattern data, delete the alarm policy;

[0118] In some embodiments, when the monitoring system determines through cluster analysis that the behavioral data of an abnormal animal matches a certain normal behavioral pattern, it considers the vehicle's current behavior to be normal. To avoid unnecessary alarms, the system will automatically delete previously generated alarm policies.

[0119] S209. If the behavioral data is not clustered by any normal behavioral pattern data, return to the step of tracking the abnormal animal and obtaining the movement rhythm data of the abnormal animal until the preset time is reached or there is no abnormal animal on the monitoring screen.

[0120] In some embodiments, when the monitoring system detects through cluster analysis that the behavioral data of an unusual animal does not match any of the preset normal behavioral pattern data, it considers the vehicle's behavior abnormal. At this point, the system will restart tracking the unusual animal, continuously collecting its movement rhythm data. During tracking, the system will continuously check whether a preset time has been reached or whether the unusual animal can still be seen in the monitoring screen. If the preset time has been reached, it means that the vehicle has been tracked for a sufficient period of time, and it is still unclear whether its behavior is illegal; tracking can then be temporarily stopped. If no unusual animal is seen in the monitoring screen, it means that the vehicle has left the monitoring range and cannot be tracked further; the current tracking process will also stop.

[0121] As can be seen, when the movement rhythm data does not conform to the preset uniformity rule, multi-dimensional behavioral data of the abnormal animal from its appearance to the present is collected. This data can comprehensively reflect its behavioral characteristics. Next, the behavioral data is clustered with preset normal behavior pattern data. Through this comparative clustering, it is possible to accurately determine whether the behavior is normal. If it is clustered with the normal pattern, it indicates that the situation is likely normal, and deleting the alarm strategy can avoid false alarms, reduce unnecessary interference, and improve alarm accuracy. If it is not clustered, new data is continuously collected, and the judgment is repeated until the conditions are met.

[0122] After step S209, the method further includes: S210, if the behavior data is not clustered by any normal behavior pattern data after a preset time has been reached or no abnormal animal is detected on the monitoring screen; the behavior data is output to the database to be determined.

[0123] The database to be determined is a database specifically designed to store behavioral data that has not yet been clearly categorized into known normal behavioral patterns. It provides a centralized storage space for further analysis and research of these special behavioral data, which helps to discover new behavioral patterns.

[0124] In some embodiments, when cluster analysis indicates that behavioral data does not match any of the preset normal behavioral patterns, the monitoring system retrieves the behavioral data from a temporary storage location and stores it in a database to be determined according to a specific format and rules. The purpose of this is to preserve this unique behavioral data so that subsequent personnel can perform more detailed analysis to explore the existence of new normal behavioral patterns.

[0125] S211. If the database to be determined is determined, add a new normal behavior pattern and determine the behavior data as sub-data of the new normal behavior pattern.

[0126] Among them, new normal behavior patterns refer to behavior patterns that have universal significance and regularity, which are discovered through the analysis and mining of special behavior data, in addition to the original preset normal behavior patterns.

[0127] In some embodiments, after further verification and confirmation by subsequent staff, these behavioral data with similar characteristics are defined as a new normal behavioral pattern. Then, the records that originally belonged to these behavioral data are marked as sub-data of the new normal behavioral pattern from the database to be determined, and the corresponding normal behavioral pattern database is updated.

[0128] As can be seen, when behavioral data is not clustered and a preset time has elapsed or there are no foreign objects in the monitoring screen, it is output to the database to be determined. This data is retained for subsequent analysis and further confirmation. Once the database to be determined is finalized, new normal behavioral patterns are added and their data is treated as sub-data. This continuously expands the range of normal behavioral patterns, making the system's understanding of normal behavior more comprehensive. Consequently, in subsequent judgments, it can more accurately distinguish between normal and abnormal behavior.

[0129] After step S213, the following steps are also included: S214, if a preset time is reached or there are no abnormal objects on the monitoring screen, and the movement rhythm data is not within the movement rhythm feature range in the illegal dumping of construction waste feature database; delete the alarm strategy.

[0130] In some embodiments, the monitoring system continuously tracks the object, acquires its movement rhythm data in real time, and compares it with the movement rhythm feature range in the illegal dumping behavior feature database. When a preset time is reached or the object is no longer visible on the monitoring screen, if the movement rhythm data is not within the feature range, it indicates that there is currently insufficient evidence to suggest that the object is illegally dumping waste. In this case, it is necessary to execute the alarm deletion strategy to avoid unnecessary alarm interference.

[0131] As can be seen, if the movement rhythm data is not within the range of the illegal dumping behavior feature database, the alarm strategy is to continuously track until the preset time or the absence of foreign objects is met, and then delete the alarm. This is to take into account the possibility that special circumstances may cause the rhythm data to be temporarily abnormal but not due to illegal dumping behavior. The alarm deletion strategy can avoid false alarms caused by such situations, reduce invalid alarm information, and allow the monitoring system to focus on the real illegal dumping behavior.

[0132] The following describes an exemplary integrated analysis system 400 for dynamic images in multiple scenes, provided by an embodiment of this application. Figure 4 This is an exemplary hardware structure diagram of an integrated analysis system 400 for dynamic images in multiple scenes provided in this application embodiment.

[0133] In some embodiments, the integrated analysis system 400 for multi-scene dynamic images is a computer device or includes a computer device. The computer device includes a processor, memory, and a network interface connected via a system bus. The processor of the computer device provides computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The database of the computer device stores data. The network interface of the computer device is used to communicate with other external terminals or servers via a network connection. In some embodiments, the network interface can be a wired network interface; in some embodiments, the network interface can also be a wireless network interface. When the computer program is executed by the processor, it implements the methods in the embodiments of this application.

[0134] Those skilled in the art will understand that Figure 4 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0135] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit it. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

[0136] As used in the above embodiments, depending on the context, the term "when..." can be interpreted as meaning "if...", "after...", "in response to determining...", or "in response to detecting...". Similarly, depending on the context, the phrase "when determining..." or "if (the stated condition or event) is interpreted as meaning "if determining...", "in response to determining...", "when (the stated condition or event) is detected", or "in response to detecting (the stated condition or event)".

[0137] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid-state drive), etc.

[0138] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. This program can be stored in a computer-readable storage medium, and when executed, it can include the processes described in the above method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as ROM or random access memory (RAM), magnetic disks, or optical disks.

[0139] Currently, a common technical method for regulating the transportation and dumping of construction waste is video surveillance using fixed cameras installed in key urban areas (such as construction site entrances and exits, and along major transportation routes). The system uses factors such as vehicle appearance, driving trajectory, and parking location to determine if illegal dumping of construction waste has occurred. For example, if a vehicle is found parking near an undesignated construction waste disposal site, or if its driving trajectory deviates from the prescribed transportation route and heads towards remote areas, it will be flagged as a suspected violation.

[0140] In practical applications, normal construction waste transportation operations involve many complex scenarios and situations. Some characteristics of normal transportation behavior are extremely similar to those of illegal dumping of construction waste, which can easily lead to misjudgments. This requires regulatory personnel to spend a lot of energy to verify these misjudgments one by one, wasting valuable human and time resources.

Claims

1. An integrated analysis method for dynamic images in multiple scenes, characterized in that, include: By deploying image acquisition devices in multiple scenarios, monitoring images of different scenarios can be obtained; The preset area of ​​the acquired monitoring image is analyzed to determine whether there are any foreign objects in the preset area. The foreign objects refer to objects in the preset area that have a preset difference threshold greater than the difference threshold between the elements in the preset area and other areas of the monitoring image. The alien body is divided into several first sub-regions, and the first sub-regions are clustered according to similarity and interval to obtain several second sub-regions; wherein the higher the similarity or the lower the interval, the higher the probability of clustering. The alien body is divided into an upper region and a lower region according to the height ratio threshold. The second sub-region inside the upper region is placed into the algorithm warehouse in order of shape complexity from large to small, and it is determined whether it has the pre-stored slag characteristics. The second sub-regions within the lower region are placed into the algorithm warehouse in order of increasing shape complexity to determine whether they have pre-stored automotive part features. When any of the second sub-regions within the upper region has pre-stored slag features and any of the second sub-regions within the lower region has pre-stored automotive part features, the alien body is tracked. During tracking, other image acquisition devices in the corresponding scene are activated to capture several target images containing the alien object. These other image acquisition devices are positioned at different locations in the corresponding scene with different shooting angles, and a volume calculation strategy is initiated. The volume calculation strategy includes: calculating the original volume information of the alien object in actual space based on the height, angle, focal length of the other image acquisition devices and the pixel size and position information occupied by the alien object in the target images; and periodically updating the volume data of the alien object at time intervals to form a volume sequence of the alien object's volume changing over time. In the event that the aforementioned alien body cannot be tracked, acquire all other alien bodies within the monitoring screen; The volume calculation strategy is applied to the other different animal bodies to obtain other volume information and other volume sequences; Two of the other volume information are selected from all the other volume information and determined as the first other volume information and the second other volume information; Determine whether the difference between the sum of the first other volume information and the second other volume information and the original volume information is less than a difference threshold; If the difference is less than the threshold, then the changing trend of the other volume sequences corresponding to the first other volume information and the second other volume information is determined; The other alien animal whose change trend is decreasing according to the first other volume information or the second other volume information is identified as the alien animal and tracked. The corresponding volume information is used to modify the original volume information.

2. The method according to claim 1, characterized in that, The difference threshold is positively correlated with the time elapsed from the loss of tracking of the alien animal to the present moment; The step of selecting two other volume information from all the other volume information to determine them as the first other volume information and the second other volume information specifically includes: Select two other volume information that are less than a preset departure threshold from all the other volume information, and determine them as the first other volume information and the second other volume information; The preset detachment threshold is positively correlated with the time elapsed from the loss of tracking of the alien body to the present moment.

3. The method according to claim 1, characterized in that, After the step of tracking the alien body, the method further includes: Generate the corresponding alarm policy; Determine whether the movement rhythm data of the alien body conforms to a preset uniformity rule; If the preset uniformity rule is met, then it is determined whether the movement rhythm data is within the movement rhythm feature range in the feature database of illegal dumping of construction waste. If the movement rhythm feature range is within the feature database of illegal dumping of construction waste, it is determined to be illegal dumping of construction waste, and the alarm strategy is activated. If the movement rhythm feature range is not within the feature database of illegal dumping of construction waste, the process returns to the step of tracking the alien body until a preset time is reached or the alien body is no longer visible on the monitoring screen.

4. The method according to claim 3, characterized in that, After the step of determining whether the movement rhythm data conforms to a preset uniformity rule, the method further includes: If the preset uniformity rule is not met, all behavioral data of the alien body from the time it appears on the monitoring screen to the current time will be collected. The behavioral data includes: driving trajectory, stopping time, speed change, and turning frequency. The behavioral data is clustered with preset normal behavioral pattern data; If the behavioral data is clustered by any of the normal behavioral pattern data, the alarm policy is deleted; If the behavioral data is not clustered by any of the normal behavioral pattern data, the step of tracking the alien animal and obtaining the movement rhythm data of the alien animal is returned until a preset time is reached or the alien animal is no longer visible on the monitoring screen.

5. The method according to claim 4, characterized in that, If the behavioral data is not clustered by any of the normal behavioral pattern data, the method returns to tracking the abnormal animal to obtain its movement rhythm data. This process continues until a preset time is reached or the abnormal animal is no longer visible on the monitoring screen. The method further includes: If the preset time is reached or the monitoring screen does not show the alien body, and the behavioral data is not clustered by any normal behavioral pattern data; The behavioral data is output to the database to be determined. If the database to be determined is determined, a new normal behavior pattern is added, and the behavior data is determined as sub-data of the new normal behavior pattern.

6. The method according to claim 3, characterized in that, If the movement rhythm feature range is not in the feature database of illegal dumping of construction waste, the step of tracking the alien body is returned until the preset time is reached or the alien body is no longer visible on the monitoring screen. If the preset time is reached, or if there are no abnormal objects on the monitoring screen, and if the movement rhythm data is not within the movement rhythm feature range in the illegal dumping of construction waste feature database, the alarm strategy will be deleted.

7. The method according to claim 1, characterized in that, The step of placing the second sub-regions within the upper region into the algorithm warehouse in descending order of shape complexity, and determining whether they possess pre-stored slag characteristics, specifically includes: Based on the image acquisition device corresponding to the second sub-region, the corresponding algorithm is selected from the algorithm repository; Based on the algorithm repository, computing resources are allocated to determine whether the soil features are pre-stored.

8. An integrated analysis system suitable for dynamic images in multiple scenes, characterized in that, The integrated analysis system for multi-scene dynamic images includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to cause the integrated analysis system for multi-scene dynamic images to perform the method as described in any one of claims 1-7.

9. A computer program product containing instructions, characterized in that, When the computer program product is run on an integrated analysis system for multi-scene dynamic images, the integrated analysis system for multi-scene dynamic images performs the method as described in any one of claims 1-7.

10. A computer-readable storage medium comprising instructions, characterized in that, When the instructions are executed on an integrated analysis system for multi-scene dynamic images, the integrated analysis system for multi-scene dynamic images performs the method as described in any one of claims 1-7.

Images