Distributed processing method and system for three-dimensional detection image
By segmenting the three-dimensional detection image into subtasks and processing it in parallel on multiple computing nodes, the problem of low processing efficiency in the traditional single computing node mode is solved, and efficient and real-time three-dimensional detection image processing is achieved.
Patent Information
- Application Number
- CN202411413732.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-11
- Publication Date
- 2025-02-25
AI Technical Summary
The traditional single computing node mode seems to be incompetent when processing high-resolution three-dimensional detection images, resulting in long processing time and low efficiency, and cannot meet the requirements of real-time and accuracy.
The distributed processing method is adopted to divide the three-dimensional detection image into multiple subtasks, and process it in parallel through multiple computing nodes, and then integrate and feedback update after generating the processing results.
The distributed processing method is significantly improved in the processing efficiency of three-dimensional image detection, solved the problem of long processing time, and met the requirements of real-time and accuracy.
Smart Images
Figure CN119515659A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of image processing, and in particular to a distributed processing method and system for three-dimensional detection images. Background Art
[0002] In today's era of rapid technological development, three-dimensional detection technology, as an important branch of computer vision and image processing, is gradually penetrating into all walks of life, from industrial automation, intelligent manufacturing to medical diagnosis, virtual reality, etc., and its application prospects are extremely broad. With the increase in the complexity of target objects and the refinement of detection requirements, the traditional single computing node processing mode has been difficult to meet the needs of efficient, real-time and accurate three-dimensional detection. Specifically, high-resolution three-dimensional detection images contain a large number of data points, which require huge computing resources for analysis and processing. The traditional single computing node mode often seems to be unable to cope with such large-scale data, resulting in long processing time and low efficiency, and cannot meet the requirements of real-time and accuracy. In addition, with the increase in data volume, data storage, transmission and synchronization have also become a major problem, further restricting the development of three-dimensional detection technology. Summary of the invention
[0003] The present application provides a distributed processing method and system for three-dimensional detection images, which solves the technical problem in the prior art that the three-dimensional detection image processing has a large amount of calculations and results in a long processing time.
[0004] In view of the above problems, the present application provides a distributed processing method and system for three-dimensional detection images.
[0005] In a first aspect of the present application, a distributed processing method for three-dimensional detection images is provided, the method comprising:
[0006] Utilize multi-eye vision imaging technology to perform three-dimensional detection of the target object to obtain a three-dimensional detection image; divide the three-dimensional detection image into multiple subtasks according to the spatial position of the target object; transmit the multiple subtasks to multiple computing nodes, and process the multiple subtasks in parallel through the multiple computing nodes to generate multiple processing results; synchronize the multiple processing results to a central computing node for interaction, and integrate them according to the interactive data set to generate a three-dimensional detection result of the target object; synchronize the three-dimensional detection result to a feedback mechanism for processing and feedback, and update the three-dimensional detection result according to the feedback result.
[0007] A second aspect of the present application provides a distributed processing system for three-dimensional detection images, the system comprising:
[0008] A three-dimensional detection module, the three-dimensional detection module is used to use multi-eye vision imaging technology to perform three-dimensional detection on the target object and obtain a three-dimensional detection image; a segmentation module, the segmentation module is used to segment the three-dimensional detection image into multiple subtasks according to the spatial position of the target object; a processing module, the processing module is used to transmit the multiple subtasks to multiple computing nodes, and the multiple subtasks are processed in parallel by the multiple computing nodes to generate multiple processing results; an integration module, the integration module is used to synchronize the multiple processing results to the central computing node for interaction, and generate a three-dimensional detection result of the target object based on the integration of the interactive data set; a feedback module, the feedback module is used to synchronize the three-dimensional detection result to the feedback mechanism for processing and feedback, and update the three-dimensional detection result according to the feedback result.
[0009] One or more technical solutions provided in this application have at least the following technical effects or advantages:
[0010] First, the target object is detected in three dimensions using multi-eye visual imaging technology to obtain a three-dimensional detection image. Then, the three-dimensional detection image is divided into multiple subtasks according to the spatial position of the target object. Further, the multiple subtasks are transmitted to multiple computing nodes, and the multiple subtasks are processed in parallel by multiple computing nodes to generate multiple processing results. Then, the multiple processing results are synchronized to the central computing node for interaction, and the three-dimensional detection results of the target object are generated based on the integration of the interactive data set. Finally, the three-dimensional detection results are synchronized to the feedback mechanism for processing and feedback, and the three-dimensional detection results are updated according to the feedback results. The technical problem of the large amount of computational processing of three-dimensional detection images in the prior art resulting in long processing time is solved, and the technical effect of improving image processing efficiency is achieved through distributed processing. BRIEF DESCRIPTION OF THE DRAWINGS
[0011] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.
[0012] Figure 1 A schematic diagram of a distributed processing method for three-dimensional detection images provided in an embodiment of the present application;
[0013] Figure 2 A schematic diagram of the structure of a distributed processing system for three-dimensional detection images provided in an embodiment of the present application.
[0014] Explanation of the reference numerals: three-dimensional detection module 11 , segmentation module 12 , processing module 13 , integration module 14 , feedback module 15 . DETAILED DESCRIPTION
[0015] The present application solves the technical problem in the prior art that the processing of three-dimensional detection images is computationally intensive and results in a long processing time by providing a distributed processing method and system for three-dimensional detection images.
[0016] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.
[0017] It should be noted that the terms "including" and "having" are intended to cover non-exclusive inclusions. For example, a process, method, system, product or server that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or modules that are not explicitly listed or are inherent to these processes, methods, products or devices.
[0018] Embodiment 1, as Figure 1 As shown, the present application provides a distributed processing method for three-dimensional detection images, wherein the method includes:
[0019] Multi-eye vision imaging technology is used to perform three-dimensional detection of target objects and obtain three-dimensional detection images.
[0020] The use of multi-camera visual imaging technology to perform three-dimensional detection of target objects aims to obtain detailed three-dimensional spatial information of the target object through the collaborative work of multiple cameras, that is, to reconstruct the three-dimensional structure of the object using image information from different perspectives to obtain a three-dimensional detection image.
[0021] Further, the method comprises:
[0022] Identify multiple acquisition angles according to the spatial structure information of the target object; use multiple image acquisition devices to collect data on the target object according to the multiple acquisition angles to obtain multi-angle images; perform feature extraction on the multi-angle images to obtain multiple feature points; traverse the multi-angle images based on the multiple feature points to perform feature matching and generate multiple matching feature point pairs; perform triangulation calculations based on the multiple matching feature point pairs to construct a three-dimensional model; use the three-dimensional model to detect and identify the target object to generate the three-dimensional detection image.
[0023] Preferably, according to the spatial structure information of the target object, including the shape, size, surface features and possible occlusion of the target object, multiple acquisition angles are set to ensure that these angles can fully cover the target object; according to the multiple acquisition angles, multiple image acquisition devices (such as high-definition cameras) are deployed to acquire images of the target object to obtain multi-angle images; each image is preprocessed, including denoising, contrast enhancement, distortion correction, etc., and the image processing algorithm is used to extract significant feature points in the image to obtain multiple feature points; in images at different angles, the feature matching algorithm is used to find corresponding feature point pairs to obtain multiple matching feature point pairs; according to the matching feature point pairs and camera parameters (such as position, posture, focal length, etc.), the triangulation principle is used to calculate the coordinates of each feature point in the three-dimensional space, and the discrete feature points are connected into a continuous three-dimensional surface through interpolation, fitting and other methods to construct a three-dimensional model of the target object; the constructed three-dimensional model is compared and analyzed with the target object, the shape, size, position, etc. of the target object are detected and identified, and a three-dimensional detection image of the target object is generated according to the detection and identification results.
[0024] The three-dimensional detection image is divided into a plurality of subtasks according to the spatial position of the target object.
[0025] In the 3D detection process, after obtaining a complete 3D detection image, in order to perform subsequent processing and analysis more efficiently, the image can be divided into multiple subtasks according to the spatial position of the target object. By breaking down complex processing tasks into multiple relatively simple and independent subtasks, parallel processing can be performed to improve overall processing efficiency. For example, according to the position and shape of the target object in 3D space, it can be divided into multiple spatial regions to ensure that each region contains a part of the target object and the overlap between regions is as small as possible.
[0026] Furthermore, the three-dimensional detection image is divided into a plurality of subtasks according to the spatial position of the target object, and the method includes:
[0027] Based on the three-dimensional model, three-dimensional coordinate parameters, object direction parameters, and object size parameters of the target object are extracted; based on the three-dimensional coordinate parameters, the object direction parameters, and the object size parameters, the spatial position of the target object is determined; according to the spatial position, a segmentation standard is determined, and the three-dimensional detection image is divided according to the segmentation standard to generate the multiple subtasks.
[0028] Preferably, three-dimensional coordinate parameters, object direction parameters, and object size parameters of the target object are extracted from the three-dimensional model, wherein the three-dimensional coordinate parameters are the coordinate information of the target object in the three-dimensional space, the object direction parameters are the direction of the target object in the three-dimensional space, and can be obtained by calculating the main axis direction of the model, a rotation matrix or a quaternion, etc., and the object size parameters are the size of the target object in the three-dimensional space, including length, width, height, etc.; based on the extracted three-dimensional coordinate parameters, object direction parameters, and object size parameters, the spatial position of the target object in the three-dimensional space can be comprehensively determined; according to the spatial position of the target object, its distribution in the three-dimensional detection image is analyzed, including the shape, size, position, and possible occlusion of the object, and based on the results of the spatial position analysis, a suitable segmentation strategy is formulated, such as dividing the target object according to its projection area in the image, or segmenting it according to different parts of the object (such as the head, body, limbs, etc.); according to the determined segmentation criteria, an image segmentation algorithm such as threshold-based segmentation, edge-based segmentation, region-based segmentation, etc. can be used to divide the three-dimensional detection image; after the division is completed, the three-dimensional detection image is divided into multiple sub-images or sub-regions, each sub-image or sub-region corresponds to a subtask.
[0029] The multiple subtasks are transmitted to multiple computing nodes, and the multiple computing nodes are used to process the multiple subtasks in parallel to generate multiple processing results.
[0030] All subtasks are organized into a task queue. According to the current load, processing capacity, available resources and other factors of the computing node, appropriate subtasks are selected from the task queue and assigned to each computing node. The assigned subtasks are transmitted to the corresponding computing node through the network or other communication mechanisms. After receiving the subtasks, each computing node processes them independently and generates multiple processing results. Through the above steps, multiple subtasks are effectively transmitted to multiple computing nodes for parallel processing, thereby significantly improving the efficiency and performance of three-dimensional detection. At the same time, through reasonable task distribution, load balancing and result collection mechanisms, the accuracy and reliability of the processing results are ensured.
[0031] Furthermore, the multiple subtasks are transmitted to multiple computing nodes, and the multiple subtasks are processed in parallel by the multiple computing nodes to generate multiple processing results. The method includes:
[0032] Analyze the task processing rates of the multiple computing nodes and construct a task allocation strategy; transmit the multiple subtasks to the multiple computing nodes according to the task allocation strategy through a data transmission protocol; determine a first computing node by random selection based on the multiple computing nodes; process the assigned subtasks in parallel with the remaining computing nodes through the first computing node to generate N task processing data, where N is an integer greater than or equal to 1; verify the N task processing data of the first computing node, output a first processing result based on the verification result, and iterate to generate the multiple processing results.
[0033] Preferably, by monitoring each computing node, the task processing rate of each computing node is obtained, including indicators such as processing time and resource utilization, and then a dynamic task allocation strategy can be constructed based on the monitoring data. The task allocation strategy can dynamically adjust the amount of tasks assigned to each node according to factors such as the current load and processing capacity of the computing node to ensure the maximization of the overall processing efficiency; according to the network environment and security requirements, a suitable data transmission protocol (such as TCP / IP, HTTP, FTP, etc.) is selected to transmit subtask data, and according to the task allocation strategy, multiple subtasks are assigned to multiple computing nodes through the data transmission protocol; a computing node is randomly selected from multiple computing nodes as the first computing node, the first computing node will receive its assigned subtask, and perform parallel calculations with the remaining computing nodes, that is, all computing nodes process their respective subtasks in parallel to generate N task processing data, where N is an integer greater than or equal to 1; the N task processing data of the first computing node are verified, including data integrity check, consistency verification, etc., and a first processing result is output according to the verification result; based on the process of the first computing node obtaining the first processing result, the same verification process is performed on the remaining computing nodes to obtain multiple processing results.
[0034] Furthermore, the task processing rates of the plurality of computing nodes are analyzed and a task allocation strategy is constructed, the method comprising:
[0035] Retrieve historical processing record data sets of the multiple computing nodes, evaluate the multiple computing nodes based on the historical processing record data sets, and generate task processing evaluation results; calculate the task processing rates of the multiple computing nodes according to the task processing evaluation results; obtain real-time network information, and construct the task allocation strategy based on the task processing rates of the multiple computing nodes combined with the real-time load data of the multiple computing nodes.
[0036] Preferably, a historical processing record data set of multiple computing nodes is retrieved from a database, and the historical processing record data set contains information such as the type of tasks, task volume, and processing time processed by each node in the past period of time; the historical processing records of the computing nodes are evaluated using methods such as statistical analysis and machine learning, and the evaluation indicators may include average processing time, processing rate (task volume / time), success rate, stability, etc. Scoring can be performed based on the evaluation indicators to obtain a task processing evaluation result for each computing node, which is used to reflect the processing capacity and performance of the computing node; based on the task processing evaluation result, the task processing rate of each computing node is calculated, such as dividing the processed task volume by the processing time to obtain the task processing rate; through a network monitoring tool, network bandwidth, delay, packet loss rate and other information are obtained in real time, and at the same time, real-time load data of the computing nodes, including CPU usage, memory occupancy, disk I / O, etc., are obtained, and the task processing rate, real-time load data and real-time network information of the computing nodes are comprehensively considered to construct a task allocation strategy, for example, by dynamically adjusting the task volume assigned to each node to ensure that tasks are evenly distributed among multiple computing nodes to avoid overloading some nodes while other nodes are idle. Through the above steps, a dynamic task allocation strategy based on computing node performance, real-time load and network information can be constructed to improve the efficiency and reliability of task processing.
[0037] The multiple processing results are synchronized to a central computing node for interaction, and the three-dimensional detection results of the target object are generated by integrating the interactive data sets.
[0038] After processing multiple subtasks in parallel and generating multiple processing results, these processing results are synchronized to the central computing node for interaction and integration to obtain the three-dimensional detection results of the target object.
[0039] Furthermore, the multiple processing results are synchronized to a central computing node for interaction, and the three-dimensional detection result of the target object is generated by integrating the interactive data set. The method includes:
[0040] The multiple processing results are synchronized to the central computing node, and the multiple processing results are traversed by the central computing node to perform feature calculation to generate multiple feature vectors; the multiple processing results are data aligned based on the multiple feature vectors to obtain multiple synchronously processed data; the multiple synchronously processed data are traversed to perform data deduplication to obtain the interactive data set.
[0041] Preferably, a computing node of the task processing rate can be selected as a central computing node, multiple processing results can be synchronized to the central computing node, multiple processing results can be traversed, feature calculation can be performed on each processing result, features of the target object can be extracted, and the extracted features can be used to form a feature vector, with each processing result corresponding to a feature vector; using the feature vectors of multiple processing results, the same or similar features in different processing results can be aligned through methods such as feature matching or similarity calculation, and further spatial or temporal alignment can be performed based on information such as the spatial position or timestamp of the processing results to ensure that the data in the multiple processing results are in the same coordinate system or timeline, and multiple synchronously processed data can be obtained after data alignment; in multiple synchronously processed data, duplicate or highly similar data can be identified and removed by comparing the similarity between the data and using methods such as hash functions. After deduplication, unique and valid processing data is retained to form a final interactive data set.
[0042] The three-dimensional detection result is synchronized to a feedback mechanism for processing and feedback, and the three-dimensional detection result is updated according to the feedback result.
[0043] The 3D detection results generated by the central computing node are synchronized to the feedback mechanism through appropriate communication protocols and channels, and the feedback is processed, including processing progress query, processing abnormal correction, and updating the 3D detection results according to the feedback results.
[0044] Furthermore, the three-dimensional detection result is synchronized to a feedback mechanism for processing and feedback, and the three-dimensional detection result is updated according to the feedback result. The method includes:
[0045] Encapsulating the three-dimensional detection result into data to generate a three-dimensional detection data packet;
[0046] Synchronize the three-dimensional detection data packet to a feedback mechanism, identify and track the three-dimensional detection data packet through the feedback mechanism, and obtain multiple tracking parameters, where the multiple tracking parameters may include a processing progress parameter and a processing exception parameter;
[0047] Generate a feedback result according to the processing progress parameter and the processing exception parameter;
[0048] An update instruction is triggered according to the feedback result, and the three-dimensional detection result is updated through the update instruction.
[0049] Preferably, the 3D detection results are organized into a standard data format, and appropriate data encapsulation technology (such as JSON, XML, binary format, etc.) is used to encapsulate the organized 3D detection results into an independent 3D detection data packet, which contains necessary information such as detection data, timestamp, and detection ID; the encapsulated 3D detection data packet is synchronized to the feedback mechanism through the network or other communication channels, and the feedback mechanism parses the received 3D detection data packet, extracts the 3D detection results and metadata therein, and identifies and tracks the 3D detection results, monitors its processing process, including tracking the processing progress, detecting anomalies that occur during the processing process, etc., and extracts multiple information from the identification and tracking process. The method can extract a tracking parameter, such as a processing progress parameter (indicating the current processing stage or completion degree) and a processing exception parameter (indicating errors or abnormal situations occurring during the processing); analyze the extracted tracking parameters, evaluate the processing status and possible problems of the 3D detection results, and generate detailed feedback results according to the processing progress parameter and the processing exception parameter. The feedback results include the current processing status (such as processing, processing completed, processing exception, etc.); trigger an update instruction according to the feedback result, and update the 3D detection results by executing the update instruction, including reprocessing the original data, correcting errors, etc.; updating the 3D detection results through the feedback mechanism can improve the accuracy of the 3D detection results.
[0050] In summary, the embodiments of the present application have at least the following technical effects:
[0051] First, the target object is detected in three dimensions using multi-eye visual imaging technology to obtain a three-dimensional detection image. Then, the three-dimensional detection image is divided into multiple subtasks according to the spatial position of the target object. Further, the multiple subtasks are transmitted to multiple computing nodes, and the multiple subtasks are processed in parallel by multiple computing nodes to generate multiple processing results. Then, the multiple processing results are synchronized to the central computing node for interaction, and the three-dimensional detection results of the target object are generated based on the integration of the interactive data set. Finally, the three-dimensional detection results are synchronized to the feedback mechanism for processing and feedback, and the three-dimensional detection results are updated according to the feedback results. The technical problem of the large amount of computational processing of three-dimensional detection images in the prior art resulting in long processing time is solved, and the technical effect of improving image processing efficiency is achieved through distributed processing.
[0052] Embodiment 2 is based on the same inventive concept as the distributed processing method for three-dimensional detection images in the above embodiment. Figure 2 As shown, the present application provides a distributed processing system for three-dimensional detection images, wherein the system includes:
[0053] A three-dimensional detection module 11, the three-dimensional detection module 11 is used to use multi-eye visual imaging technology to perform three-dimensional detection on the target object and obtain a three-dimensional detection image; a segmentation module 12, the segmentation module 12 is used to segment the three-dimensional detection image into multiple subtasks according to the spatial position of the target object; a processing module 13, the processing module 13 is used to transmit the multiple subtasks to multiple computing nodes, and the multiple subtasks are processed in parallel by the multiple computing nodes to generate multiple processing results; an integration module 14, the integration module 14 is used to synchronize the multiple processing results to the central computing node for interaction, and generate a three-dimensional detection result of the target object by integrating the interactive data set; a feedback module 15, the feedback module 15 is used to synchronize the three-dimensional detection result to the feedback mechanism for processing and feedback, and update the three-dimensional detection result according to the feedback result.
[0054] Furthermore, the three-dimensional detection module 11 is used to perform the following method:
[0055] Identify multiple acquisition angles according to the spatial structure information of the target object; use multiple image acquisition devices to collect data on the target object according to the multiple acquisition angles to obtain multi-angle images; perform feature extraction on the multi-angle images to obtain multiple feature points; traverse the multi-angle images based on the multiple feature points to perform feature matching and generate multiple matching feature point pairs; perform triangulation calculations based on the multiple matching feature point pairs to construct a three-dimensional model; use the three-dimensional model to detect and identify the target object to generate the three-dimensional detection image.
[0056] Furthermore, the segmentation module 12 is used to perform the following method:
[0057] Based on the three-dimensional model, three-dimensional coordinate parameters, object direction parameters, and object size parameters of the target object are extracted; based on the three-dimensional coordinate parameters, the object direction parameters, and the object size parameters, the spatial position of the target object is determined; according to the spatial position, a segmentation standard is determined, and the three-dimensional detection image is divided according to the segmentation standard to generate the multiple subtasks.
[0058] Furthermore, the processing module 13 is used to execute the following method:
[0059] Analyze the task processing rates of the multiple computing nodes and construct a task allocation strategy; transmit the multiple subtasks to the multiple computing nodes according to the task allocation strategy through a data transmission protocol; determine a first computing node by random selection based on the multiple computing nodes; process the assigned subtasks in parallel with the remaining computing nodes through the first computing node to generate N task processing data, where N is an integer greater than or equal to 1; verify the N task processing data of the first computing node, output a first processing result based on the verification result, and iterate to generate the multiple processing results.
[0060] Furthermore, the processing module 13 is used to execute the following method:
[0061] Retrieve historical processing record data sets of the multiple computing nodes, evaluate the multiple computing nodes based on the historical processing record data sets, and generate task processing evaluation results; calculate the task processing rates of the multiple computing nodes according to the task processing evaluation results; obtain real-time network information, and construct the task allocation strategy based on the task processing rates of the multiple computing nodes combined with the real-time load data of the multiple computing nodes.
[0062] Furthermore, the integration module 14 is used to execute the following method:
[0063] The multiple processing results are synchronized to the central computing node, and the multiple processing results are traversed by the central computing node to perform feature calculation to generate multiple feature vectors; the multiple processing results are data aligned based on the multiple feature vectors to obtain multiple synchronously processed data; the multiple synchronously processed data are traversed to perform data deduplication to obtain the interactive data set.
[0064] Furthermore, the feedback module 15 is used to perform the following method:
[0065] The three-dimensional detection result is data-encapsulated to generate a three-dimensional detection data packet; the three-dimensional detection data packet is synchronized to a feedback mechanism, and the three-dimensional detection data packet is identified and tracked through the feedback mechanism to obtain multiple tracking parameters, and the multiple tracking parameters may include processing progress parameters and processing exception parameters; feedback results are generated according to the processing progress parameters and in combination with the processing exception parameters; an update instruction is triggered according to the feedback result, and the three-dimensional detection result is detected and updated through the update instruction.
[0066] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. And the above-mentioned specific embodiments of this specification are described. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
[0067] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.
[0068] This specification and drawings are merely exemplary illustrations of the present application and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these modifications and variations.
Claims
1. A distributed processing method for three-dimensional detection images, characterized in that: The method comprises: Use multi-eye vision imaging technology to perform three-dimensional detection of target objects and obtain three-dimensional detection images; Segmenting the three-dimensional detection image into a plurality of subtasks according to the spatial position of the target object; Transmitting the multiple subtasks to multiple computing nodes, and performing parallel processing on the multiple subtasks by the multiple computing nodes to generate multiple processing results; Synchronizing the multiple processing results to a central computing node for interaction, and integrating the interactive data sets to generate a three-dimensional detection result of the target object; The three-dimensional detection result is synchronized to a feedback mechanism for processing and feedback, and the three-dimensional detection result is updated according to the feedback result.
2. A distributed processing method for three-dimensional detection images according to claim 1, characterized in that: The method comprises: Identify multiple acquisition angles according to the spatial structure information of the target object; Using multiple image acquisition devices to acquire data of the target object according to the multiple acquisition angles to obtain multi-angle images; Performing feature extraction on the multi-angle image to obtain multiple feature points; Traversing the multi-angle images to perform feature matching based on the multiple feature points to generate multiple matching feature point pairs; Perform triangulation calculation based on the multiple matching feature point pairs to construct a three-dimensional model; The target object is detected and identified using the three-dimensional model to generate the three-dimensional detection image.
3. A distributed processing method for three-dimensional detection images according to claim 2, characterized in that: The three-dimensional detection image is divided into a plurality of subtasks according to the spatial position of the target object, and the method includes: Extracting three-dimensional coordinate parameters, object direction parameters, and object size parameters of the target object based on the three-dimensional model; Determine the spatial position of the target object based on the three-dimensional coordinate parameters, the object direction parameters, and the object size parameters; A segmentation standard is determined according to the spatial position, and the three-dimensional detection image is divided according to the segmentation standard to generate the multiple subtasks.
4. A distributed processing method for three-dimensional detection images according to claim 1, characterized in that: The multiple subtasks are transmitted to multiple computing nodes, and the multiple subtasks are processed in parallel by the multiple computing nodes to generate multiple processing results. The method includes: Analyze the task processing rates of the multiple computing nodes and build a task allocation strategy; Transmitting the plurality of subtasks to the plurality of computing nodes according to the task allocation strategy through a data transmission protocol; Determine a first computing node by performing random selection based on the multiple computing nodes; Processing the assigned subtask in parallel with the remaining computing nodes through the first computing node to generate N task processing data, where N is an integer greater than or equal to 1; The N task processing data of the first computing node are verified, and a first processing result is output according to the verification result, thereby iterating to generate the multiple processing results.
5. A distributed processing method for three-dimensional detection images as claimed in claim 4, characterized in that: Analyzing the task processing rates of the plurality of computing nodes and constructing a task allocation strategy, the method includes: Retrieving historical processing record data sets of the plurality of computing nodes, evaluating the plurality of computing nodes based on the historical processing record data sets, and generating task processing evaluation results; Calculating the task processing rates of the plurality of computing nodes according to the task processing evaluation results; Real-time network information is obtained, and the task allocation strategy is constructed based on the task processing rates of the multiple computing nodes combined with the real-time load data of the multiple computing nodes.
6. A distributed processing method for three-dimensional detection images according to claim 1, characterized in that: The multiple processing results are synchronized to a central computing node for interaction, and a three-dimensional detection result of the target object is generated by integrating the interactive data set. The method includes: Synchronize the multiple processing results to the central computing node, and perform feature calculation on the multiple processing results through the central computing node to generate multiple feature vectors; Performing data alignment on the multiple processing results based on the multiple feature vectors to obtain multiple synchronously processed data; The plurality of synchronously processed data are traversed to perform data deduplication to obtain the interactive data set.
7. A distributed processing method for three-dimensional detection images according to claim 1, characterized in that: The three-dimensional detection result is synchronized to a feedback mechanism for processing and feedback, and the three-dimensional detection result is updated according to the feedback result. The method includes: Encapsulating the three-dimensional detection result into data to generate a three-dimensional detection data packet; Synchronize the three-dimensional detection data packet to a feedback mechanism, identify and track the three-dimensional detection data packet through the feedback mechanism, and obtain multiple tracking parameters, where the multiple tracking parameters may include a processing progress parameter and a processing exception parameter; Generate a feedback result according to the processing progress parameter and the processing exception parameter; An update instruction is triggered according to the feedback result, and the three-dimensional detection result is updated through the update instruction.
8. A distributed processing system for three-dimensional detection images, characterized in that: A distributed processing method for three-dimensional detection images according to any one of claims 1 to 7, the system comprising: A three-dimensional detection module, which is used to perform three-dimensional detection of a target object using multi-eye visual imaging technology to obtain a three-dimensional detection image; A segmentation module, the segmentation module is used to segment the three-dimensional detection image into multiple subtasks according to the spatial position of the target object; A processing module, the processing module is used to transmit the multiple subtasks to multiple computing nodes, and process the multiple subtasks in parallel through the multiple computing nodes to generate multiple processing results; An integration module, the integration module is used to synchronize the multiple processing results to a central computing node for interaction, and integrate the interactive data sets to generate a three-dimensional detection result of the target object; A feedback module is used to synchronize the three-dimensional detection result to a feedback mechanism for processing and feedback, and update the three-dimensional detection result according to the feedback result.