Remote data monitoring and processing method and device

Abstract

The present application provides a remote data monitoring and processing method and device, which obtains a monitoring image and shooting parameters collected by a remote camera device, performs pixel depth calculation processing on the monitoring image, and obtains depth monitoring information corresponding to the monitoring image; performs object recognition processing on the monitoring image to determine a first object area and a first object extension area; performs color quality evaluation processing on each pixel in the monitoring image according to the color parameters of each pixel and shooting scene information to obtain a color quality evaluation index for each pixel; identifies target pixels whose quality evaluation index is lower than a preset threshold, and constructs a non-significant change area according to the identified target pixels; determines a first depth equalization value and a second depth equalization value of the non-significant change area according to the depth information; performs accuracy verification on the depth monitoring information; if the verification passes, generates a remote monitoring video according to the monitoring image and the depth monitoring information.

Description

Technical Field

[0001] The embodiments of the present application relate to the fields of cloud computing and video technology, and in particular to a remote data monitoring and processing method and device. Background Art

[0002] With the rapid development of information technology, remote monitoring systems have become an important part of security, environmental monitoring and other fields. Remote monitoring systems collect monitoring images through remote camera equipment and transmit them to remote monitoring platforms for analysis and processing to achieve real-time monitoring of specific areas or objects.

[0003] Currently, in video surveillance, it is necessary to add depth to the acquired surveillance images to generate remote surveillance videos with visual effects. However, due to various factors such as complex scenes in the surveillance images, the acquisition of depth information is not accurate, resulting in poor video quality of the generated remote surveillance videos. Summary of the invention

[0004] The embodiments of the present application provide a remote data monitoring and processing method and device, which can improve the accuracy and effect of remote monitoring video generation.

[0005] The present application provides a remote data monitoring and processing method, which is applicable to a remote monitoring system, wherein the remote monitoring system includes a remote cloud monitoring platform and a remote camera device. The monitoring and processing method is applicable to the remote cloud monitoring platform, and includes:

[0006] Acquire a monitoring image and shooting parameters collected by the remote camera device, perform pixel depth calculation processing on the monitoring image, and obtain depth monitoring information corresponding to the monitoring image, wherein the monitoring image includes a first object and a second object, the first object and the second object have an association relationship, the depth monitoring information includes a depth value corresponding to each pixel in the monitoring image, and the shooting parameters are shooting parameters of the remote camera device;

[0007] Performing object recognition processing on the monitoring image to determine a first object area and a first object extension area corresponding to the first object area, wherein the first object area includes all pixels of the first monitoring object;

[0008] Performing color quality evaluation processing on each pixel in the monitoring image according to the color parameter of each pixel and the shooting scene information to obtain a color quality evaluation index for each pixel, wherein the color quality evaluation index is used to measure the color information of the pixel;

[0009] Identify target pixels whose quality evaluation index is lower than a preset threshold between the first object region and the first object extension region, and construct a non-significant change region according to the identified target pixels; the non-significant change region is a region where the transition color change between the region boundary and the image background presents a non-significant change state;

[0010] Determine, according to the depth information, a first depth equalization value of the non-significantly changed area and a second depth equalization value of the second object area;

[0011] Performing accuracy verification on the depth monitoring information according to the shooting parameters, the association relationship, the first depth balance value, and the second depth balance value;

[0012] If the verification passes, a remote monitoring video is generated based on the monitoring image and the deep monitoring information.

[0013] Another aspect of the present application embodiment provides a remote data monitoring and processing device, which is applicable to a remote cloud monitoring platform and includes:

[0014] Another aspect of the present application provides a computer-readable storage medium, in which instructions are stored. When the computer-readable storage medium is run on a computer, the computer is enabled to execute the above-mentioned methods.

[0015] An embodiment of the present application provides a remote data monitoring and processing method, which obtains a monitoring image and shooting parameters collected by the remote camera device, performs pixel depth calculation processing on the monitoring image, and obtains depth monitoring information corresponding to the monitoring image, wherein the monitoring image includes a first object and a second object, the first object and the second object have an associated relationship, the depth monitoring information includes a depth value corresponding to each pixel in the monitoring image, and the shooting parameters are the shooting parameters of the remote camera device; performs object recognition processing on the monitoring image to determine a first object area and a first object extension area corresponding to the first object area, wherein the first object area includes all pixels of the first monitoring object; and performs pixel depth calculation processing on each pixel in the monitoring image according to the color parameter of each pixel and the shooting scene information. Color quality evaluation processing, obtaining a color quality evaluation index for each pixel, the color quality evaluation index is used to measure the color information of the pixel; identifying a target pixel whose quality evaluation index is lower than a preset threshold between the first object area and the first object extension area, and constructing a non-significant change area according to the identified target pixel; the non-significant change area is an area where the transition color change between the area edge and the image background presents a non-significant change state; determining a first depth balance value of the non-significant change area and a second depth balance value of the second object area according to the depth information; checking the accuracy of the depth monitoring information according to the shooting parameters, the association relationship, the first depth balance value and the second depth balance value; if the check is passed, generating a remote monitoring video according to the monitoring image and the depth monitoring information. The scheme can identify the non-significant change area on the monitoring image, and check the accuracy of the depth information of the monitoring image based on the shooting parameters of the remote camera device, the association relationship between the first and second objects in the image, the depth balance value of the non-significant change area and the depth balance value of the second object area. When the check is passed, generating a remote monitoring video based on the monitoring image information and the depth monitoring information can ensure the accuracy of the depth information of the monitoring image, thereby improving the accuracy and effect of remote monitoring video generation. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a schematic diagram of the architecture of the remote data monitoring and processing system according to an embodiment of the present application;

[0017] Figure 2 This is a flow chart of an embodiment of a remote data monitoring and processing method in an embodiment of the present application;

[0018] Figure 3 is a flowchart of an embodiment of three-dimensional reconstruction in an embodiment of the present application;

[0019] Figure 4It is a schematic diagram of the structure of a remote data monitoring and processing device in an embodiment of the present application;

[0020] Figure 5 It is a schematic diagram of an embodiment of a computer device in an embodiment of the present application. DETAILED DESCRIPTION

[0021] The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of the present application and the accompanying drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequential order. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in other orders. In addition, the terms "including" and "corresponding to" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0022] With the rapid development of information technology, cloud technology has become an indispensable part of our daily life. Cloud technology covers various network technologies, information technology, integration technology, management platform technology and application technology based on the cloud computing business model. It can build a resource pool to achieve on-demand allocation and flexible and convenient use. Cloud computing technology will become an important support for the backend services of technical network systems, especially in scenarios that require a large amount of computing and storage resources, such as video websites, image websites and portal websites.

[0023] With the rapid development and widespread application of the Internet industry, each item may have its own identification mark in the future, and this information needs to be transmitted to the background system for logical processing. Different levels of data will be processed separately. Data from all walks of life need a strong system backing to support it, and cloud computing technology is the key to achieving this goal.

[0024] Cloud security is the sum of security software, hardware, users, institutions and security cloud platforms based on the cloud computing business model. It combines emerging technologies and concepts such as parallel processing, grid computing and unknown virus behavior judgment. Through a mesh structure composed of a large number of clients, cloud security can monitor abnormal software behavior in the network, collect the latest information of Trojans and malicious programs on the Internet, and send it to the server for automatic analysis and processing. Subsequently, the solutions for viruses and Trojans are distributed to each client.

[0025] Currently, in the field of remote monitoring, due to possible occlusion or complex geometric structure scenes in the monitoring image, the visual system faces challenges in depth estimation, which may cause errors in the algorithm's analysis of image content, thereby affecting the accuracy of object recognition. This inaccuracy will make the depth data obtained from the visual system unreliable, which in turn affects the quality of the video generated based on this data, making the visual effect unsatisfactory.

[0026] The remote data monitoring and processing method and device provided in the embodiments of the present application can apply cloud computing, cloud security and other technologies to achieve the generation of remote monitoring videos, and can more efficiently and safely generate remote monitoring videos with high-quality images.

[0027] In one embodiment, the present application proposes a remote monitoring method, which is applied to Figure 1 The remote monitoring system shown in the reference Figure 1 , Figure 1 FIG. 1 is a schematic diagram of the architecture of the remote monitoring system in the embodiment of the present application. Figure 1 As shown, the system includes: a remote cloud monitoring platform and at least one remote camera device, wherein the remote cloud monitoring platform and the remote camera device are connected via a network.

[0028] Among them, the cloud monitoring platform can be an independent cloud server or a cloud server cluster or distributed system composed of multiple cloud servers. The cloud server can be a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (CDNs), as well as big data and artificial intelligence platforms.

[0029] Among them, the remote camera device can collect image data of the monitored objects in the monitoring area and upload it to the remote cloud monitoring platform.

[0030] The remote cloud monitoring platform can obtain the monitoring image and shooting parameters collected by the remote camera device, perform pixel depth calculation processing on the monitoring image, and obtain depth monitoring information corresponding to the monitoring image, wherein the monitoring image includes a first object and a second object, the first object and the second object have an association relationship, the depth monitoring information includes the depth value corresponding to each pixel in the monitoring image, and the shooting parameters are the shooting parameters of the remote camera device. The monitoring image is subjected to object recognition processing to determine a first object area and a first object extension area corresponding to the first object area, wherein the first object area includes all pixels of the first monitoring object; color quality evaluation processing is performed on each pixel in the monitoring image according to the color parameter of each pixel and the shooting scene information to obtain a color quality evaluation index of each pixel, wherein the color quality evaluation index is used to measure the color information of the pixel; target pixels whose quality evaluation index is lower than a preset threshold are identified between the first object area and the first object extension area, and a non-significant change area is constructed according to the identified target pixels; the non-significant change area is an area where the transition color change between the area boundary and the image background presents a non-significant change state; a first depth balance value of the non-significant change area and a second depth balance value of the second object area are determined according to the depth information; the depth monitoring information is verified for accuracy according to the shooting parameters, the association relationship, the first depth balance value and the second depth balance value; if the verification is passed, a remote monitoring video is generated according to the monitoring image and the depth monitoring information. The remote cloud monitoring platform can send the generated remote monitoring video to the terminal device for display and playback.

[0031] The embodiment of the present application can identify non-significantly changed areas on a surveillance image, and verify the accuracy of the depth information of the surveillance image based on the shooting parameters of the remote camera device, the correlation relationship between the first and second objects in the image, the depth balance value of the non-significantly changed area, and the depth balance value of the second object area. When the verification passes, a remote surveillance video is generated based on the non-significantly changed area, the surveillance image information, and the depth monitoring information, which can ensure the accuracy of the depth information of the surveillance image, thereby improving the accuracy and effect of remote surveillance video generation.

[0032] In order to solve the above problems, the present application proposes a remote data monitoring and processing method, which is generally executed by a remote cloud monitoring platform such as a cloud server. The remote data monitoring and processing device can be set in the remote cloud monitoring platform.

[0033] The following will introduce the monitoring and processing method in this application in detail from the perspective of the remote data monitoring and processing device. Please refer to Figure 2 In one embodiment, the monitoring processing method includes:

[0034] S101. Acquire a monitoring image and shooting parameters collected by the remote camera device, perform pixel depth calculation processing on the monitoring image, and obtain depth monitoring information corresponding to the monitoring image.

[0035] The monitoring image includes a first object and a second object, the first object and the second object are associated with each other, and the depth monitoring information includes a depth value corresponding to each pixel in the monitoring image and a shooting parameter of the camera device.

[0036] In one embodiment, remote camera devices can be deployed in multiple monitoring areas respectively, with one remote camera device deployed in each monitoring area to collect monitoring image data of objects in the monitoring area and upload them to the remote cloud monitoring platform in real time, so that the remote cloud monitoring platform can generate and output remote monitoring videos based on the monitoring image data.

[0037] The monitoring image may specifically include an image of at least one object, which may be a living body such as a person or an animal, or a static object, etc. In one embodiment, the monitoring image may include a first object and a second object, for example, two objects or two persons, which is not specifically limited here.

[0038] Among them, the depth monitoring information may include the depth information of the monitoring image, for example, it may be depth image information, etc. In one embodiment, the remote cloud monitoring platform may perform pixel depth calculation processing on the monitoring image to obtain the depth information corresponding to the monitoring image. In one embodiment, if the depth information is expressed as a depth image, the depth image is an image obtained by performing depth estimation on the monitoring image, and there is a one-to-one correspondence between the pixels in the monitoring image and the pixels in the depth image. In one embodiment, the depth calculation can form a depth image by calculating the distance between the object and the shooting point, which can also be called an optical flow map, wherein the depth image contains the parallax of the object between different images. The parallax is used to obtain the depth information of the target object, that is, the distance between the target object and the shooting point.

[0039] Parallax is a key indicator to measure the distance between the target and the shooting point, which reveals the depth information of the target. In some contexts, parallax is also called optical flow. To understand this concept more intuitively, imagine a passenger sitting in a moving car, staring at the scene outside the car window. As the car moves, the passenger will find that the distant scenery changes slowly, while the nearby objects pass by quickly. Imagine that the passenger captures two pictures at two consecutive moments. If the two pictures contain the same scenery, the pixel coordinate differences of these scenes in the two pictures are usually inconsistent. Specifically, if the pixel coordinates of a scene in the two pictures are significantly different, it can be inferred that the scene is closer to the shooting point; conversely, if the pixel coordinates of a scene in the two pictures are slightly different, then the scene is relatively far away from the shooting point. In short, scenes that are closer will produce larger parallax, while scenes that are farther away will produce smaller parallax. Therefore, the difference in pixel coordinates of the same scene in different pictures can be regarded as parallax.

[0040] In one embodiment, the depth monitoring information may also include shooting parameters of the remote camera device, which refer to the parameters used by the remote camera device when capturing images, and may include: shooting angle, lighting conditions, focal length, color stability, saturation, exposure, etc.

[0041] In one embodiment, the shooting angle of the camera may be the field of view (FOV) of the camera lens, which is a parameter describing the viewing angle range that the camera lens can cover. The size of the field of view determines the range of scenes that the camera can "see", including the horizontal field of view, vertical field of view, diagonal field of view, etc.

[0042] In one embodiment, the remote device head device can collect monitoring images of the monitoring area, and then upload the monitoring images and the shooting parameters of the device to the remote cloud monitoring platform for data processing. Among them, the first object and the second object are different objects in the same shooting scene, which are represented by people or objects. For example, different objects in the same monitoring image, for example, two objects in a monitoring image. In one embodiment, the first object and the second object are two different objects with an association relationship, wherein the association relationship can be set according to actual needs. For example, in one embodiment, the association relationship can be that the first object and the second object are the same object type, such as both are people, or object types. For another example, in one embodiment, the association relationship can be a position association relationship, that is, the first object and the second object are two objects with associated positions in the monitoring image. Specifically, in one embodiment, the position distance between the first object and the second object in the monitoring image is less than a preset distance threshold, such as two objects with adjacent positions.

[0043] S102: Perform object recognition processing on the monitoring image to determine a first object area and a first object extension area corresponding to the first object area, wherein the first object area includes all pixels of the first monitoring object.

[0044] In one embodiment, after acquiring the surveillance image, object recognition processing can be performed on the surveillance image to obtain position information of the objects (first and second objects) in the surveillance image, and the object area of ​​the first object in the surveillance image (i.e., the first object area) is determined based on the position information.

[0045] Among them, the first object area can be specifically expressed as a rectangular frame that includes the first object in the monitoring image, and the rectangular frame includes various pixel points of the first object, which can be used to mark the position of the first object on the monitoring image (such as two-dimensional and three-dimensional coordinate values). In one embodiment, it can also include information such as the attributes of the marked first object (such as the object name).

[0046] In one embodiment, during the actual recognition process, the boundary of the first object area happens to be divided into a complex color area with significant color changes, which may cause a split between the first object area and the image background information. This embodiment can identify the extended area range corresponding to the first object area to determine the area where cracks may be formed and then perform corresponding processing to avoid a sense of split between the boundary of the first object and the image background information in the remote monitoring video generated later.

[0047] The first object extended area is an area obtained by extending the position of the first object area. For example, a rectangular area obtained by extending the rectangular area where the first object is located outward by a certain distance is the first object extended area.

[0048] In one embodiment, performing object recognition processing on the surveillance image to determine the first object area and the object extension area corresponding to the first object area may specifically include:

[0049] Performing object recognition processing on the surveillance image to obtain a first object area and position information of the first object area in the surveillance image;

[0050] A first object extension area corresponding to the first object area is determined in the monitoring image based on the position information and the position offset information, wherein the position offset information indicates position information of area boundary pixels of the first object area that need to be offset.

[0051] The position offset information may be set according to actual needs, such as a distance of 100 pixels.

[0052] For example, in one embodiment, the monitoring image is subjected to object recognition processing, and the remote cloud monitoring platform can use advanced computer vision technology, such as convolutional neural network (CNN) in deep learning, to identify the first object area in the image and obtain the precise location information of the area in the image. The location information generally includes parameters such as the center coordinates, width, and height of the object area, which help to accurately locate the position of the object in the image.

[0053] Then, the remote cloud monitoring platform determines the first object extension area corresponding to the first object area based on the acquired position information and the position offset information. The position offset information refers to the offset required on the boundary pixel points of the object area. These offsets can be preset fixed values ​​or dynamically calculated values ​​based on specific application scenarios. By applying these offsets on the boundaries of the object area, the remote cloud monitoring platform can expand the object area, thereby obtaining a wider range of contextual information, which helps to more comprehensively analyze the object and its surrounding environment.

[0054] In the embodiment of the present application, the determination of the extended area may also adopt a simple geometric method, such as adding a fixed number of pixels on each boundary of the object area, or adopt a more complex algorithm, such as an adaptive expansion method based on image content.

[0055] Through the above steps, the object area and its corresponding extended area in the monitoring image can be effectively determined, providing richer information for subsequent image analysis and processing.

[0056] In one embodiment, in order to improve the accuracy of the object area and thus improve the effect and accuracy of the remote monitoring video, a window sliding method can be used to select a rectangular frame containing each pixel point of the first object from the monitoring image to form a first object area.

[0057] Specifically, sliding windows of different sizes are applied to the surveillance image in various directions (left-right, up-down). At each window position, a pre-trained classifier, such as a support vector machine (SVM) classifier, is executed. If the classification probability obtained by the classifier in the current window is higher than a preset threshold, then it can be considered that the target object is detected in the window. This probability threshold is set according to the specific application scenario and requirements, and there is no fixed standard. Afterwards, the optimal window is selected as the object area of ​​the first object using techniques such as non-maximum suppression (NMS). In addition to sliding windows, algorithms such as selective search (SS), edge box (EB) or region generation network (RPN) can also be used to extract object areas.

[0058] In one embodiment, after the first object region is detected, an extended region of the first object region may be formed by expanding the region in proportion to a preset magnification factor, and the specific method of obtaining the first object extended region is not limited.

[0059] S103. Perform color quality evaluation processing on each pixel in the monitoring image according to the color parameter of each pixel and the shooting scene information to obtain a color quality evaluation index for each pixel, where the quality evaluation index is used to measure the color information of the pixel.

[0060] The color parameter of a pixel is a parameter that characterizes the color of the pixel, such as an RGB value, an HSV value, or other color space representation of the pixel.

[0061] Among them, the shooting scene information is the scene of the monitoring image collected by the remote camera device. The specific scene type can be divided according to actual needs. For example, in one embodiment, it can be divided into indoor monitoring scenes, outdoor monitoring scenes, community monitoring scenes, public place monitoring scenes, traffic monitoring scenes, etc. In one embodiment, it can also be divided into standard definition camera scenes, high definition camera scenes, etc. according to signal clarity. In one embodiment, it can also be divided according to purpose, including industrial-grade camera scenes and household-grade cameras.

[0062] In the embodiment of the present application, the color quality score is performed on each pixel in the monitoring image. The color quality evaluation index can reflect the color change, saturation, brightness change, etc. of each pixel in the monitoring image, and then the pixel points with significant color change, saturation, and brightness change can be obtained, so that the quality of the object in the image can be better identified.

[0063] In one embodiment, color quality evaluation processing is performed on each pixel in the surveillance image according to the color parameter of each pixel and the shooting scene information, and the color quality evaluation index of each pixel may include:

[0064] Performing color space conversion on the color parameters of each pixel to obtain converted color parameters;

[0065] Correct the converted color parameters of each pixel according to the shooting scene; analyze the color saturation and brightness of each pixel according to the corrected color parameters (too high or too low saturation and brightness may affect color quality);

[0066] Evaluate the color consistency of adjacent pixels in the monitoring image according to the corrected color parameters to obtain a consistency evaluation result;

[0067] The color quality evaluation index of each pixel is calculated according to the consistency evaluation result, color saturation and brightness.

[0068] The calculation method of color quality evaluation index can be rule-based or learned through machine learning models. For example, support vector machines (SVM) or neural networks can be used to predict the color quality of each pixel, and these models can be trained based on a large amount of image data with color quality annotated.

[0069] In the embodiment of the present application, the color quality evaluation index calculation method can be specifically as follows:

[0070] In order to analyze color properties more accurately, the color of the pixel is first converted from the RGB color space to a color space more suitable for evaluation, such as LAB or HSV. The LAB color space is more commonly used in color evaluation because it is closer to human visual perception.

[0071] According to the shooting scene information, such as the lighting conditions corresponding to different scenes, the image is color corrected to eliminate the color deviation caused by environmental factors. Common white balance algorithms include the gray world assumption and the perfect reflection assumption. The gray world assumes that the average color of the image is gray, and white balance is achieved by adjusting the gain of the RGB channels. The perfect reflection assumes that the brightest point in the image is a perfectly reflected white, and the color is corrected by adjusting the gain. These algorithms can effectively correct the color deviation caused by changes in lighting conditions and improve the color quality of the image.

[0072] Evaluate the color consistency of adjacent pixels in the image. You can use local variance or gradient methods to detect areas where the color changes are too drastic. Specifically: calculate the local variance or gradient of each pixel; if the local variance or gradient exceeds a certain threshold, it is considered that the color changes in this area are drastic and the color quality is poor. Through this analysis, you can identify areas in the image where the color changes are abnormal, providing a basis for further color quality evaluation.

[0073] Specifically, to evaluate the color consistency of adjacent pixels in an image, local variance or gradient methods can be used to detect areas where the color changes are too drastic. Taking local variance as an example, the specific steps are as follows:

[0074] A. Calculate the local variance of each pixel:

[0075] For each pixel, the pixel value variance of the pixels in its area (3*3pix) is calculated to obtain the local variance of each pixel.

[0076] B. Calculate color consistency information such as color consistency score based on the local variance of each pixel.

[0077] For each pixel, if its local variance is greater than the preset threshold, it indicates that the color changes in the domain area are drastic and the color quality is poor. At this time, the color consistency score of the pixel can be set, for example, a fixed value such as 0 (which can be set according to actual needs). If the local variance of the pixel is not greater than the preset threshold, it indicates that the color changes in the domain area are not obvious and the color quality is good. The color consistency score of the pixel can be set to a fixed value such as 100 or 1. The color consistency score (Y) of each pixel can be obtained in the above manner.

[0078] In one embodiment, the color saturation and brightness of each pixel can be analyzed to evaluate the color quality. Too high or too low saturation and brightness may affect the color quality. The S (saturation) and V (brightness) values ​​in the HSV color space can be used for analysis:

[0079]

[0080] If the saturation and brightness are beyond the preset reasonable range, the color quality is considered poor. Through this analysis, the color attributes of each pixel can be quantified.

[0081] Based on the above analysis results, a color quality evaluation index is calculated for each pixel. This index can be a comprehensive score or a set of values ​​reflecting different color attributes. The comprehensive score can be calculated using the following formula:

[0082] Color quality score = w1×saturation S+w2×brightness V+w3×color consistency score Y.

[0083] Among them, w1, w2, and w3 are weights that can be adjusted according to actual application requirements. Through this comprehensive evaluation, the color quality evaluation index of each pixel can be obtained, thereby comprehensively evaluating the color quality of the image.

[0084] S104, identifying target pixels whose quality evaluation index is lower than a preset threshold between the first object area and the first object extension area, and constructing a non-significant change area based on the identified target pixels; the non-significant change area is an area where the transition color change between the area boundary and the image background presents a non-significant change state. In one embodiment, after obtaining the quality evaluation index of each pixel, the remote cloud monitoring platform can identify target pixels whose quality evaluation index is lower than a preset threshold between the first object area and the object extension area. If the color quality evaluation index of the pixel is greater than the preset threshold, it indicates that the color change or color change of the pixel is not obvious, and it can be regarded as part of the non-significant area. Finally, the non-significant change area is determined based on the target pixel. For example, the area surrounded by target pixels below the preset threshold can be regarded as the non-significant change area.

[0085] Among them, the quality score indicators of the pixels in the non-significant change area can all be lower than the preset threshold, or a certain proportion of the pixels can be lower than the preset threshold, which can be set according to actual needs.

[0086] In this embodiment, after the first object area is acquired, the monitoring image has a corresponding relationship with the pixel points in the depth information, and the depth information of the pixel points in the depth information is positively correlated with the color information of the pixel points. Therefore, a color quality evaluation is performed on each pixel point on the monitoring image to obtain a color quality evaluation index for each pixel point. The color changes of each pixel point, such as brightness and saturation changes, can be reflected through the color quality evaluation index of each pixel point.

[0087] Then, the target pixel points whose color quality evaluation index is lower than a preset threshold between the first object area and the first object extension area can be detected to search for the non-significant change area of ​​the image color. In this way, a remote monitoring video can be subsequently generated based on the non-significant change area to reduce the obvious significant visual changes between the boundary of the object in the video and the image background, thereby improving the visual display effect of the remote monitoring video.

[0088] Among them, the second object area can be an image area in the monitoring image that contains the second object, which can be specifically expressed as a rectangular frame that contains the second object in the monitoring image. The rectangular frame includes various pixel points of the second object and can be used to mark the position of the first object on the monitoring image (such as two-dimensional and three-dimensional coordinate values). In one embodiment, it can also include information such as the attributes of the marked first object (such as the object name).

[0089] In an embodiment of the present application, the second object area acquisition method and the second object area method can be based on object recognition detection of the surveillance image, for example, detecting the position information of the second object in the surveillance image, and determining the image area where the second object is located based on the position information.

[0090] S105 : Determine a first depth equalization value for the non-significantly changed area and a second depth equalization value for the second object area according to the depth information.

[0091] The first depth equalization value may be an average depth value of pixels in the non-significantly changed area, and the second depth equalization value may be an average depth value of pixels in the second object area.

[0092] In one embodiment, when the depth information is expressed as a depth image, the pixels in the monitoring image and the depth image have a corresponding relationship in position, so the corresponding pixels in the non-salient area and the depth values ​​corresponding to the pixels can be found in the depth information such as the depth image, and then, based on the depth values ​​of the corresponding pixels in the depth image, the average depth value of the pixels in the non-salient area, i.e., the first depth balance value, can be calculated. Similarly, the depth value information of the corresponding pixels and pixels in the second object area can be found in the depth image, and the average depth value of the pixels in the second object area, i.e., the second depth balance value, can be calculated based on the found depth value information.

[0093] S106: Verify the accuracy of the depth monitoring information according to the shooting parameters, the association relationship, the first depth balance value, and the second depth balance value.

[0094] In an embodiment of the present application, after obtaining the first depth balance value and the second depth balance value of the second object area, the remote cloud monitoring platform can verify the accuracy of the depth monitoring information according to the shooting parameters, the association relationship, the first depth balance value and the second depth balance value. If the verification passes, a remote monitoring video is generated according to the non-significant change area, the monitoring image and the depth monitoring information. In this way, a remote monitoring video with accurate image can be generated, thereby improving the quality and visual effect of the remote monitoring video.

[0095] In one embodiment, there are multiple ways to verify the accuracy of the depth monitoring information according to the shooting parameters, the association relationship, the first depth balance value, and the second depth balance value. In one embodiment, in order to improve the accuracy of the depth monitoring information verification, it may include:

[0096] fusing the first depth equalization value and the second depth equalization value according to the shooting parameter and the association relationship to obtain a fused depth equalization value;

[0097] If the fused depth balance value is within the preset balance value range, it is determined that the accuracy check of the depth monitoring information has passed.

[0098] In this embodiment of the present application, the first depth equalization value and the second depth equalization value are obtained. Since the depth value can represent the actual distance between the object and the shooting point, relative to the background information, the first object can be understood as the foreground information of the monitoring image. In the actual scene, relative to the distance between the background information and the shooting point, the foreground information should be closer to the shooting point. Therefore, the first depth equalization value and the second depth equalization value of the non-significantly changed area of ​​the first object are used separately for fusion processing.

[0099] In actual shooting scenarios, due to the existence of factors such as parallax, the depth information of the monitoring image in the monitoring scene is not accurately judged. Therefore, the embodiment of the present application takes into account the relationship between the associated objects in the image and the shooting parameters of the device in the monitoring scene, and fuses the depth values ​​of the associated objects. The accuracy of the depth information of the monitoring image is judged based on the fused depth value, thereby improving the accuracy of the depth information.

[0100] In one embodiment, according to the shooting parameter and the association relationship, the first depth equalization value and the second depth equalization value are fused to obtain a fused depth equalization value, including:

[0101] respectively setting a first depth fusion weight of the first object area and a second depth fusion weight of the second object area according to the shooting parameters and the association relationship;

[0102] The first depth equalization value and the second depth equalization value are weighted according to the first depth fusion weight and the second depth fusion weight to obtain a fused depth equalization value.

[0103] In one embodiment, if the association relationship includes that the distance between the first object and the second object is less than a preset distance threshold, then the target distance range within which the distance between the first object and the second object lies can be obtained;

[0104] Acquire, from a preset range weight mapping relationship set according to the target distance range, a first preset fusion weight of the first object area and a second preset fusion weight corresponding to the second object area;

[0105] The first preset fusion weight and the second preset fusion weight are adjusted according to the shooting parameters to obtain a first depth fusion weight of the first object area and a second depth fusion weight of the second object area.

[0106] Among them, the range weight mapping relationship set can be a pre-set mapping relationship set, which includes the mapping relationship between the sample distance range and the sample fusion weight, and can be expressed as a mapping table. In the embodiment of the present application, the target distance range of the distance between the two association relationships, i.e., the first object and the second object, can be determined first, and then the weight corresponding to the target distance range can be found in the mapping relationship set to obtain the first preset fusion weight of the first object area and the second preset fusion weight corresponding to the second object area.

[0107] Since different distance ranges between two related objects in the same surveillance image will result in different degrees of depth information judgment, in an embodiment of the present application, preset fusion weights corresponding to different distance ranges can be pre-set. In this way, the corresponding preset fusion weight can be obtained when the distance range between the first and second objects is obtained.

[0108] Similarly, in the embodiments of the present application, the influence of different shooting parameters such as shooting angle, lighting conditions, etc. on the acquisition of depth information will also be considered. Therefore, in order to accurately determine whether the depth information of the monitoring image in the current scene is accurate, it is necessary to adjust or correct the first preset fusion weight and the second preset fusion weight according to the shooting parameters, so as to improve the verification accuracy of the depth information.

[0109] There are various adjustment parameters for the fusion weight according to the shooting parameters. For example, an adjustment ratio can be set according to different shooting parameters.

[0110] In one embodiment, considering that the shooting angle has a high degree of influence on the verification accuracy of the depth information, priority is given to the case where the shooting parameters include the shooting angle of the remote camera device. Specifically, the first preset fusion weight and the second preset fusion weight are adjusted according to the shooting parameters to obtain the first depth fusion weight of the first object area and the second depth fusion weight of the second object area, including:

[0111] According to the shooting angle, a weight adjustment coefficient corresponding to the shooting angle is obtained from a preset coefficient set, wherein the coefficient set includes a mapping relationship between preset shooting angles and preset adjustment coefficients, and the preset adjustment system includes preset depth weight adjustment coefficients of a first object and a second object;

[0112] The first preset fusion weight and the second preset fusion weight are adjusted according to the weight adjustment coefficient to obtain a first depth fusion weight of the first object area and a second depth fusion weight of the second object area.

[0113] In the embodiment of the present application, considering the influence of the shooting angle and the verification of the depth information, the association between various shooting angles and the adjustment coefficients can be preset, and then after the shooting angle is obtained, the corresponding adjustment coefficient can be found based on the shooting angle.

[0114] S107: If the verification passes, generate a remote monitoring video according to the monitoring image and the deep monitoring information.

[0115] After the deep monitoring information of the monitoring image is verified, the remote cloud monitoring platform can generate a remote monitoring video according to the non-significant change area, the monitoring image and the deep monitoring information.

[0116] like Figure 2 As shown, generating a remote monitoring video according to the monitoring image and the deep monitoring information may include:

[0117] S201, constructing a three-dimensional image block according to the monitoring image and the depth monitoring information;

[0118] S202, generating a camera motion trajectory based on each frame of sample image in the remote monitoring video sample, wherein the camera motion trajectory is used to indicate the camera position and orientation information corresponding to each frame of template image taken by the camera;

[0119] S203: Perform image rendering according to the three-dimensional image block and the camera motion trajectory to generate a remote monitoring video.

[0120] Among them, the 3D image block, also known as the 3D mesh image, is a high-precision 3D mesh obtained by a 3D reconstruction algorithm based on the surveillance image and depth information. A 3D mesh is a data structure used to represent the geometric structure of a 3D object or scene. It consists of vertices, edges, and faces, which together define the shape and surface of an object or scene.

[0121] The camera motion trajectory refers to the position and orientation information of the camera that changes over time during the shooting process, usually represented by a series of (time, position, orientation) tuples. The position information defines the coordinates of the camera in three-dimensional space, while the orientation information describes the direction and orientation of the camera through Euler angles or quaternions. This information can be manually designed, automatically tracked, or generated based on preset templates, and is widely used in computer vision, animation, virtual reality, augmented reality and other fields to achieve tasks such as three-dimensional reconstruction, target tracking, motion analysis, and visual effect enhancement. In an embodiment of the present application, the camera motion trajectory may include the camera position and orientation information corresponding to each frame of sample image taken by the camera, and may also include camera posture information such as the camera rotation angle. The remote monitoring video sample template is a video uploaded to the platform by the monitoring personnel or manufacturers, which indicates the style of the remote monitoring video, such as the video including the video playback timeline and the video playback style.

[0122] In the embodiment of the present application, after the depth information is verified, a video can be rendered based on the three-dimensional image blocks and the camera motion trajectory. This solution can obtain accurate three-dimensional image blocks corresponding to the monitoring image and depth information through three-dimensional scene reconstruction, and then can generate high-quality remote monitoring video based on the three-dimensional image blocks and high-quality camera motion trajectories, thereby improving the visual effect of the remote monitoring video.

[0123] In one embodiment, in order to improve the visual effect of the remote monitoring video, Figure 3As shown, the depth monitoring information is a depth image, and performing three-dimensional reconstruction according to the monitoring image and the depth monitoring information to obtain a three-dimensional image block may include:

[0124] S301. Perform resolution operation processing on the depth image to obtain a processed monitoring image and a processed depth image.

[0125] A resolution reduction operation is performed on the input monitoring image and depth image to generate image versions with corresponding reduced details, that is, simplified versions of the monitoring image and depth image. For example, this can be achieved through a downsampling operation at a certain rate, which can reduce the amount of data and improve efficiency.

[0126] S302: Perform geometric mapping processing according to the spatial position information of each pixel in the processed depth image to construct a basic three-dimensional image block.

[0127] By using the spatial positioning of each pixel in the simplified version of the depth image and geometric mapping technology, a basic 3D model framework, i.e., a basic 3D image block, is constructed. This step is the key conversion process of converting depth data into a 3D space grid.

[0128] S303: Perform detail enhancement processing on the basic three-dimensional image block to obtain a transitional three-dimensional image block.

[0129] The basic 3D model framework is subjected to detail enhancement processing to restore texture details reduced due to resolution reduction, thereby obtaining a transitional 3D image block with richer details.

[0130] S304: Perform color mapping on the transitional three-dimensional image block according to color parameters of pixels in the monitoring image to obtain a three-dimensional image block.

[0131] Based on the color data of the monitoring image, the transitional 3D model is color mapped to make the model more realistic visually, and finally form a complete 3D model. The whole process extracts and constructs a 3D model with color information from the 2D image data through a series of operations.

[0132] In one embodiment, before performing video rendering based on the surveillance image and the depth image, the playback time of the remote surveillance video sample can be obtained, and the camera motion rate corresponding to each frame sample image in the remote surveillance video sample can be obtained based on the playback time of the remote surveillance video sample, and the camera motion rate can be interpolated to better simulate the camera's motion rate, and then the camera motion trajectory can be accurately obtained, so that subsequent rendering can obtain high-quality remote surveillance video.

[0133] From the above, it can be seen that the embodiment of the present application provides a remote data monitoring and processing method that can identify non-significantly changed areas on the monitoring image, and based on the shooting parameters of the remote camera device, the correlation relationship between the first and second objects in the image, the depth balance value of the non-significantly changed area and the depth balance value of the second object area, the accuracy of the depth information of the monitoring image is verified. When the verification passes, a remote monitoring video is generated based on the non-significantly changed area, the monitoring image information and the depth monitoring information, which can ensure the accuracy of the depth information of the monitoring image, thereby improving the accuracy and effect of the remote monitoring video generation.

[0134] Furthermore, the present application can also generate remote monitoring videos based on non-significantly changed areas. Taking into account the color transition of the object area, the problem of color separation or mismatch between the monitored object and the entire image in the remote monitoring video can be reduced, thereby improving the display effect of the entire remote monitoring video.

[0135] In one embodiment, based on the above remote data monitoring and processing method, a remote data monitoring and processing device is also provided. The monitoring and processing device can be integrated in a remote cloud monitoring platform. Figure 4 , which is a remote data monitoring and processing device, specifically includes:

[0136] An acquisition unit 401 is used to acquire a monitoring image and shooting parameters collected by the remote camera device, perform pixel depth calculation processing on the monitoring image, and obtain depth monitoring information corresponding to the monitoring image, wherein the monitoring image includes a first object and a second object, the first object and the second object have an association relationship, the depth monitoring information includes a depth value corresponding to each pixel in the monitoring image, and the shooting parameters are shooting parameters of the remote camera device;

[0137] an identification unit 402, configured to perform object identification processing on the monitoring image to determine a first object region and a first object extension region corresponding to the first object region, wherein the first object region includes all pixels of the first monitoring object;

[0138] An evaluation index unit 403 is used to perform color quality evaluation processing on each pixel in the monitoring image according to the color parameter of each pixel and the shooting scene information, so as to obtain a color quality evaluation index of each pixel, wherein the color quality evaluation index is used to measure the color information of the pixel;

[0139] The region detection unit 404 is used to identify target pixel points whose quality evaluation index is lower than a preset threshold between the first object region and the first object extension region, and construct a non-significant change region according to the identified target pixel points; the non-significant change region is a region where the transition color change between the region edge and the image background presents a non-significant change state;

[0140] A depth calculation unit 405 is used to determine a first depth equalization value of the non-significantly changed area and a second depth equalization value of the second object area according to the depth information;

[0141] A verification unit 406, configured to verify the accuracy of the depth monitoring information according to the shooting parameters, the association relationship, the first depth balance value, and the second depth balance value;

[0142] The remote video unit 407 is used to generate a remote monitoring video according to the monitoring image and the deep monitoring information if the verification is passed.

[0143] In one embodiment, the inspection unit 406 is used to:

[0144] fusing the first depth equalization value and the second depth equalization value according to the shooting parameter and the association relationship to obtain a fused depth equalization value;

[0145] If the fused depth balance value is greater than the preset balance value, it is determined that the accuracy check of the depth monitoring information has passed.

[0146] In one embodiment, the checking unit 406 is configured to: respectively set a first depth fusion weight of the first object area and a second depth fusion weight of the second object area according to the shooting parameter and the association relationship;

[0147] The first depth equalization value and the second depth equalization value are weighted according to the first depth fusion weight and the second depth fusion weight to obtain a fused depth equalization value.

[0148] In one embodiment, the association relationship includes that the distance between the first object and the second object is less than a preset distance threshold; the verification unit 406 is used to:

[0149] If the distance between the first object and the second object is less than a preset distance threshold, obtaining a target distance range within which the distance between the first object and the second object lies;

[0150] Acquire, from a preset range weight mapping relationship set according to the target distance range, a first preset fusion weight of the first object area and a second preset fusion weight corresponding to the second object area;

[0151] The first preset fusion weight and the second preset fusion weight are adjusted according to the shooting parameters to obtain a first depth fusion weight of the first object area and a second depth fusion weight of the second object area.

[0152] In one embodiment, the shooting parameters at least include a shooting angle of the remote camera device; the inspection unit 406 is used to:

[0153] According to the shooting angle, a weight adjustment coefficient corresponding to the shooting angle is obtained from a preset coefficient set, wherein the coefficient set includes a mapping relationship between preset shooting angles and preset adjustment coefficients, and the preset adjustment system includes preset depth weight adjustment coefficients of a first object and a second object;

[0154] The first preset fusion weight and the second preset fusion weight are adjusted according to the weight adjustment coefficient to obtain a first depth fusion weight of the first object area and a second depth fusion weight of the second object area.

[0155] In one embodiment, the identification unit 402 is configured to:

[0156] Performing object recognition processing on the surveillance image to obtain a first object area and position information of the first object area in the surveillance image;

[0157] A first object extension area corresponding to the first object area is determined in the monitoring image based on the position information and the position offset information, wherein the position offset information indicates position information of area boundary pixels of the first object area that need to be offset.

[0158] In one embodiment, the evaluation index unit 403 is used to:

[0159] Performing color space conversion on the color parameters of each pixel to obtain converted color parameters;

[0160] Correcting the converted color parameters of each pixel according to the shooting scene; analyzing the color saturation and brightness of each pixel according to the corrected color parameters;

[0161] Evaluate the color consistency of adjacent pixels in the monitoring image according to the corrected color parameters to obtain a consistency evaluation result;

[0162] The color quality evaluation index of each pixel is calculated according to the consistency evaluation result, color saturation and brightness.

[0163] In one embodiment, the remote video unit 407 is specifically configured to:

[0164] Performing three-dimensional reconstruction according to the non-significantly changed area, the monitoring image and the depth monitoring information to obtain a three-dimensional image block;

[0165] Generate the camera motion trajectory of the video based on each frame sample image in the remote monitoring video sample;

[0166] Image rendering is performed according to the three-dimensional image block and the camera motion trajectory to generate a remote monitoring video.

[0167] In one embodiment, the depth monitoring information is a depth image, and the remote video unit 407 is specifically used to:

[0168] A resolution operation is performed on the depth image to obtain a processed monitoring image and a processed depth image.

[0169] A geometric mapping process is performed according to the spatial position information of each pixel in the processed depth image to construct a basic three-dimensional image block.

[0170] Perform detail enhancement processing on the basic three-dimensional image block to obtain a transitional three-dimensional image block.

[0171] According to the color parameters of the pixels in the monitoring image, color mapping is performed on the transitional three-dimensional image block to obtain a three-dimensional image block.

[0172] In the embodiments of the present application, the computer device can be configured as a terminal or a computer device. When the computer device is configured as a terminal, the terminal serves as the execution subject to implement the technical solution provided in the embodiments of the present application; when the computer device is configured as a computer device, the computer device serves as the execution subject to implement the technical solution provided in the embodiments of the present application; or, the technical solution provided in the present application is implemented through interaction between the terminal and the computer device, which is not limited in the embodiments of the present application.

[0173] Figure 5 : is a schematic diagram of the structure of a computer device provided in an embodiment of the present application. The computer device 500 may have relatively large differences due to different configurations or performances, and may include one or more central processors 521 and memories 532, and one or more storage media for storing application programs or data (e.g., one or more mass storage devices). Among them, the memory 532 and the storage medium may be temporary storage or permanent storage. The program stored in the storage medium may include one or more modules (not shown in the figure), and each module may include a series of instruction operations in the computer device. Furthermore, the central processor 521 may be configured to communicate with the storage medium and execute a series of instruction operations in the storage medium on the computer device 500.

[0174] The steps performed by the computer device in the above embodiment can be based on the Figure 5 The computer device structure shown in the figure. Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the above-described systems, devices and units can refer to the corresponding processes in the aforementioned method embodiments, and will not be repeated here.

[0175] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0176] The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. In addition, each functional unit in each embodiment of the present application may be integrated into a processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated units may be implemented in the form of hardware or in the form of software functional units.

[0177] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk and other media that can store program code.

Claims

1. A remote data monitoring and processing method, characterized in that: Applicable to a remote monitoring system, the remote monitoring system includes a remote cloud monitoring platform and a remote camera device, and the monitoring processing method is applicable to the remote cloud monitoring platform, including: Acquire a monitoring image and shooting parameters collected by the remote camera device, perform pixel depth calculation processing on the monitoring image, and obtain depth monitoring information corresponding to the monitoring image, wherein the monitoring image includes a first object and a second object, the first object and the second object have an association relationship, the depth monitoring information includes a depth value corresponding to each pixel in the monitoring image, and the shooting parameters are shooting parameters of the remote camera device; Performing object recognition processing on the monitoring image to determine a first object area and a first object extension area corresponding to the first object area, wherein the first object area includes all pixels of the first monitoring object; Performing color quality evaluation processing on each pixel in the monitoring image according to the color parameter of each pixel and the shooting scene information to obtain a color quality evaluation index for each pixel, wherein the color quality evaluation index is used to measure the color information of the pixel; Identify target pixels whose quality evaluation index is lower than a preset threshold between the first object region and the first object extension region, and construct a non-significant change region according to the identified target pixels; the non-significant change region is a region where the transition color change between the region boundary and the image background presents a non-significant change state; Determine a first depth equalization value for the non-significantly changed area and a second depth equalization value for the second object area according to the depth monitoring information; Performing accuracy verification on the depth monitoring information according to the shooting parameters, the association relationship, the first depth balance value, and the second depth balance value; If the verification passes, a remote monitoring video is generated based on the monitoring image and the deep monitoring information.

2. The method according to claim 1, characterized in that The accuracy of the depth monitoring information is verified according to the shooting parameter, the association relationship, the first depth balance value, and the second depth balance value, including: fusing the first depth equalization value and the second depth equalization value according to the shooting parameter and the association relationship to obtain a fused depth equalization value; If the fused depth balance value is greater than the preset balance value, it is determined that the accuracy check of the depth monitoring information has passed.

3. The method according to claim 2, characterized in that According to the shooting parameter and the association relationship, the first depth equalization value and the second depth equalization value are fused to obtain a fused depth equalization value, including: respectively setting a first depth fusion weight of the first object area and a second depth fusion weight of the second object area according to the shooting parameters and the association relationship; The first depth equalization value and the second depth equalization value are weighted according to the first depth fusion weight and the second depth fusion weight to obtain a fused depth equalization value.

4. The method according to claim 3, characterized in that The association relationship includes that the distance between the first object and the second object is less than a preset distance threshold; According to the shooting parameters and the association relationship, respectively setting a first depth fusion weight of the first object area and a second depth fusion weight of the second object area includes: If the distance between the first object and the second object is less than a preset distance threshold, obtaining a target distance range within which the distance between the first object and the second object lies; Acquire, from a preset range weight mapping relationship set according to the target distance range, a first preset fusion weight of the first object area and a second preset fusion weight corresponding to the second object area; The first preset fusion weight and the second preset fusion weight are adjusted according to the shooting parameters to obtain a first depth fusion weight of the first object area and a second depth fusion weight of the second object area.

5. The method according to claim 4, characterized in that The shooting parameters at least include the shooting angle of the remote camera device; The first preset fusion weight and the second preset fusion weight are adjusted according to the shooting parameter to obtain a first depth fusion weight of the first object area and a second depth fusion weight of the second object area, including: According to the shooting angle, a weight adjustment coefficient corresponding to the shooting angle is obtained from a preset coefficient set, wherein the coefficient set includes a mapping relationship between preset shooting angles and preset adjustment coefficients, and the preset adjustment coefficients include preset depth weight adjustment coefficients of a first object and a second object; The first preset fusion weight and the second preset fusion weight are adjusted according to the weight adjustment coefficient to obtain a first depth fusion weight of the first object area and a second depth fusion weight of the second object area.

6. The method according to any one of claims 1 to 5, characterized in that: Performing object recognition processing on the surveillance image to determine a first object area and an object extension area corresponding to the first object area includes: Performing object recognition processing on the surveillance image to obtain a first object area and position information of the first object area in the surveillance image; A first object extension area corresponding to the first object area is determined in the monitoring image based on the position information and the position offset information, wherein the position offset information indicates position information of area boundary pixels of the first object area that need to be offset.

7. The method according to claim 6, characterized in that According to the color parameters of each pixel and the shooting scene information, a color quality evaluation process is performed on each pixel in the monitoring image to obtain a color quality evaluation index of each pixel, including: Performing color space conversion on the color parameters of each pixel to obtain converted color parameters; Correcting the converted color parameters of each pixel according to the shooting scene; analyzing the color saturation and brightness of each pixel according to the corrected color parameters; Evaluate the color consistency of adjacent pixels in the monitoring image according to the corrected color parameters to obtain a consistency evaluation result; The color quality evaluation index of each pixel is calculated according to the consistency evaluation result, color saturation and brightness.

8. The method according to claim 7, characterized in that The generating of the remote monitoring video according to the monitoring image and the deep monitoring information comprises: Performing three-dimensional reconstruction according to the monitoring image and the depth monitoring information to obtain a three-dimensional image block; Generate the camera motion trajectory of the video based on each frame sample image in the remote monitoring video sample; Image rendering is performed according to the three-dimensional image block and the camera motion trajectory to generate a remote monitoring video.

9. The method according to claim 8, characterized in that The depth monitoring information is a depth image, and three-dimensional reconstruction is performed according to the monitoring image and the depth monitoring information to obtain a three-dimensional image block, including: Performing resolution operation processing on the monitoring image and the depth image respectively to obtain a processed monitoring image and a processed depth image; Performing geometric mapping processing according to the spatial position information of each pixel in the processed depth image to construct a basic three-dimensional image block; Performing detail enhancement processing on the basic three-dimensional image block to obtain a transitional three-dimensional image block; According to the color parameters of the pixels in the processed monitoring image, color mapping is performed on the transitional three-dimensional image block to obtain a three-dimensional image block.

10. A remote data monitoring and processing device, the monitoring and processing device is suitable for a remote cloud monitoring platform, characterized in that: include: an acquisition unit, configured to acquire a monitoring image and shooting parameters collected by a remote camera device, perform pixel depth calculation processing on the monitoring image, and obtain depth monitoring information corresponding to the monitoring image, wherein the monitoring image includes a first object and a second object, the first object and the second object have an associated relationship, the depth monitoring information includes a depth value corresponding to each pixel in the monitoring image, and the shooting parameters are shooting parameters of the remote camera device; an identification unit, configured to perform object identification processing on the monitoring image to determine a first object area and a first object extension area corresponding to the first object area, wherein the first object area includes all pixels of the first monitoring object; An evaluation index unit, used to perform color quality evaluation processing on each pixel in the monitoring image according to the color parameter of each pixel and the shooting scene information, to obtain a color quality evaluation index of each pixel, wherein the color quality evaluation index is used to measure the color information of the pixel; A region detection unit, configured to identify target pixel points whose quality evaluation index is lower than a preset threshold between the first object region and the first object extension region, and construct a non-significant change region according to the identified target pixel points; the non-significant change region is a region where the transition color change between the region edge and the image background presents a non-significant change state; A depth calculation unit, configured to determine a first depth equalization value of the non-significantly changed area and a second depth equalization value of the second object area according to the depth monitoring information; a verification unit, configured to verify the accuracy of the depth monitoring information according to the shooting parameters, the association relationship, the first depth balance value, and the second depth balance value; The remote video unit is used to generate a remote monitoring video according to the monitoring image and the deep monitoring information if the verification is passed.

Images