Artificial intelligence data acquisition method and device, equipment and storage medium

Abstract

The invention provides an artificial intelligence data acquisition method and device, equipment and a storage medium. The artificial intelligence data acquisition method comprises the following steps: acquiring a frame picture in a target field scene by utilizing an acquisition module of an edge equipment terminal deployed in the target field scene; preprocessing the collected frame picture in the target field scene by using a processing module in the edge device terminal to obtain a preprocessed frame picture; and transmitting the preprocessed frame picture to a corresponding acquisition platform by using a transmission module in the edge device terminal, so that the acquisition platform analyzes and processes the preprocessed frame picture to obtain a data acquisition result. According to the method, the data transmission quantity is greatly reduced in a mode of collecting the frame pictures, so that the project cost and the flow consumption are reduced.

Description

Technical Field

[0001] The present invention relates to the field of data collection technology, and specifically, to an artificial intelligence data collection method, device, equipment and storage medium, and more specifically, to a low-cost and high-efficiency artificial intelligence data collection method, device, equipment and storage medium for field scenes. Background Art

[0002] Data collection and monitoring in the wild is a challenging task. Traditional manual inspection operations are time-consuming, labor-intensive and inefficient, so it has become a trend to introduce artificial intelligence to replace manual inspection operations. However, there are some difficulties in collecting data in open fields, such as lack of network coverage and large traffic consumption. In addition, for sample engineers, the operation of manually viewing videos and intercepting samples is also very cumbersome, resulting in a long sample collection time and reduced work efficiency.

[0003] There are already some solutions that attempt to solve the problem of data collection in the wild. Among them, using 4G / 5G networks for communication is a way to solve the network coverage problem, but due to the characteristics of open fields, the network cannot cover it, resulting in ineffective data transmission. Secondly, due to the large amount of traffic generated by video transmission, the cost cannot be controlled.

[0004] Therefore, a low-cost and efficient artificial intelligence data collection method is needed to solve the above problems. Summary of the invention

[0005] The main purpose of the present invention is to solve the problem in the prior art that when collecting data in field scenes, effective data transmission cannot be performed due to lack of network coverage, and the large amount of traffic generated by video transmission and the uncontrollable cost.

[0006] The first aspect of the present invention provides an artificial intelligence data collection method, including: using a collection module of an edge device terminal deployed in a target field scene to collect frame images in the target field scene; using a processing module in the edge device terminal to preprocess the collected frame images in the target field scene to obtain preprocessed frame images; using a transmission module in the edge device terminal to transmit the preprocessed frame images to a corresponding collection platform, so that the collection platform analyzes and processes the preprocessed frame images to obtain data collection results.

[0007] Optionally, in a first implementation manner of the first aspect of the present invention, before using the acquisition module of the edge device terminal deployed in the target field scene to collect frame images in the target field scene, the method also includes: providing multiple edge devices; setting a corresponding edge device terminal topology structure according to the target field scene, and deploying multiple edge device terminals in the target field scene according to the set topology structure; performing acquisition settings on the multiple edge device terminals after deployment; wherein the multiple edge device terminals after the acquisition settings are used to collect frame images in the target field scene.

[0008] Optionally, in a second implementation of the first aspect of the present invention, the step of performing collection settings on the multiple deployed edge device terminals includes: Set the corresponding acquisition parameters for each of the multiple edge device terminals after deployment, and configure the corresponding image push addresses for the multiple edge device terminals; wherein the acquisition parameters include at least one of the acquisition time period and the acquisition interval frame rate; the image push address is used to transmit the frame image to the corresponding acquisition platform; accordingly, the acquisition module of the edge device terminal deployed in the target field scene is used to collect the frame image in the target field scene, including: based on the acquisition parameters corresponding to each of the multiple edge device terminals, collecting the frame image in the target field scene; accordingly, the transmission module in the edge device terminal is used to transmit the pre-processed frame image to the corresponding acquisition platform, including: using the transmission module in the edge device terminal to transmit the pre-processed frame image to the corresponding acquisition platform based on the image push address.

[0009] Optionally, in a third implementation method of the first aspect of the present invention, the processing module in the edge device terminal is used to preprocess the collected frame images in the target wild scene to obtain the preprocessed frame images, including: using the processing module in the edge device terminal to deduplicate the collected frame images in the target wild scene; using the processing module in the edge device terminal to filter the deduplicated frame images according to preset requirements to obtain processed frame images.

[0010] Optionally, in a fourth implementation of the first aspect of the present invention, the using a processing module in the edge device terminal to perform deduplication processing on the frame images in the collected target outdoor scene includes: using a convolutional neural network to extract feature vectors of the collected continuous frame images, calculating the similarity of the continuous frame images based on the feature vectors of the continuous frame images, and when the similarity is higher than a preset value, performing deduplication processing; in, Indicates the nth frame of the picture; represents the n+1th frame of the picture; represents the dot product of the vector; Represents the magnitude of a vector.

[0011] Optionally, in a fifth implementation method of the first aspect of the present invention, the preprocessed frame image is transmitted to a corresponding acquisition platform by using a transmission module in the edge device terminal, so that the acquisition platform analyzes and processes the preprocessed frame image to obtain a data acquisition result, including: the preprocessed frame image is transmitted to a corresponding acquisition platform by using a transmission module in the edge device terminal, so that the analysis module in the acquisition platform extracts features of the preprocessed frame image, generates data acquisition results based on corresponding extracted features, and enables the acquisition platform to display the data acquisition results.

[0012] The second aspect of the present invention provides an artificial intelligence data acquisition device, including: an edge device terminal acquisition module, used to collect frame images in a target field scene; an edge device terminal processing module, used to preprocess the collected frame images in the target field scene to obtain preprocessed frame images; an edge device terminal transmission module, used to transmit the preprocessed frame images to a corresponding acquisition platform; an acquisition platform receiving and storage module, used to receive and save the preprocessed frame images.

[0013] Optionally, in a first implementation of the second aspect of the present invention, a corresponding edge device terminal topology structure is set according to a target outdoor scene, and multiple edge device terminals are deployed in the target outdoor scene with the set topology structure; The multiple edge device terminals after deployment are set up for collection; wherein the multiple edge device terminals after collection setting are used to collect frame images in the target outdoor scene.

[0014] Optionally, in a second implementation of the second aspect of the present invention, the acquisition settings for the multiple edge device terminals after deployment include: setting acquisition parameters corresponding to each of the multiple edge device terminals after deployment, and configuring the multiple edge device terminals with respective corresponding picture push addresses; wherein the acquisition parameters include at least one of the acquisition time period and the acquisition interval frame rate; the picture push address is used to transmit the frame picture to the corresponding acquisition platform; Accordingly, the edge device terminal acquisition module is used to acquire frame images in the target outdoor scene based on acquisition parameters corresponding to each of the multiple edge device terminals; Correspondingly, the edge device terminal transmission module is used to transmit the pre-processed frame image to the corresponding acquisition platform based on the image push address.

[0015] Optionally, in a third implementation of the second aspect of the present invention, the edge device terminal processing module includes: a filter for filtering the collected frame images in the target outdoor scene to meet preset requirements, and obtain a filtered frame image. The filtering of the collected frame images in the target outdoor scene to meet preset requirements includes: at least one of deduplication processing and filtering processing to obtain the target image.

[0016] Optionally, in a fourth implementation of the second aspect of the present invention, the deduplication processing is performed on the frame images in the collected target outdoor scene using a processing module in the edge device terminal, including: extracting feature vectors of the collected continuous frame images using a convolutional neural network, calculating the similarity of the continuous frame images based on the feature vectors of the continuous frame images, and performing deduplication processing when the similarity is higher than a preset value; in, Indicates the nth frame of the picture; represents the n+1th frame of the picture; represents the dot product of the vector; Represents the magnitude of a vector.

[0017] Optionally, in a fifth implementation of the second aspect of the present invention, an analysis module within the acquisition platform is used to extract features from the preprocessed frame images, generate data acquisition results based on corresponding extracted features, and display the data acquisition results.

[0018] The third aspect of the present invention provides an electronic device, comprising: a memory and at least one processor, wherein instructions are stored in the memory, and the memory and the at least one processor are interconnected via lines; the at least one processor calls the instructions in the memory so that the electronic device executes the above-mentioned artificial intelligence data collection method.

[0019] A fourth aspect of the present invention provides a computer-readable storage medium, in which instructions are stored. When the computer-readable storage medium is run on a computer, the computer executes the above-mentioned artificial intelligence data collection method.

[0020] Compared with the prior art, the present invention has the following beneficial effects: 1. Reduce project costs and save traffic: The traditional sample collection method uses 4G / 5G to transmit video, which causes huge traffic and uncontrollable costs. However, the present invention greatly reduces the amount of data transmission by collecting frame pictures, thereby reducing project costs and traffic consumption; 2. Reduce the workload of sample engineers and improve work efficiency: The traditional sample collection method requires sample engineers to view the video frame by frame and intercept and save samples, which is complicated and time-consuming. However, the present invention greatly reduces the workload of sample engineers and improves work efficiency by automatically collecting frame images and sending them to the collection platform in real time. 3. Improve the accuracy of samples: The present invention can filter out some unnecessary content according to its own needs, thereby improving the accuracy of samples; compared with traditional sample collection methods, it can better meet the requirements for collected data; 4. In field scenarios, the artificial intelligence data collection method provided by the present invention has the advantages of reducing costs, saving traffic, reducing workload, improving work efficiency and improving sample accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments made with reference to the following drawings: Figure 1 A first flow chart of the artificial intelligence data collection method provided in an embodiment of the present invention.

[0022] Figure 2 A second flow chart of the artificial intelligence data collection method provided in an embodiment of the present invention.

[0023] Figure 3 A third flow chart of the artificial intelligence data collection method provided in an embodiment of the present invention.

[0024] Figure 4 A fourth flow chart of the artificial intelligence data collection method provided in an embodiment of the present invention.

[0025] Figure 5 A schematic diagram of the structure of an artificial intelligence data acquisition device provided in an embodiment of the present invention.

[0026] Figure 6 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0027] The embodiment of the present invention provides an artificial intelligence data collection method, device, equipment and storage medium, which uses multiple edge device terminals deployed in the target field scene to collect frame images in the target field scene, and pre-processes the collected frame images in the target field scene, and transmits the pre-processed frame images to the corresponding collection platform, and the collection platform stores the received frame images. The present invention solves the problem that when collecting data in the field scene, effective data transmission cannot be performed due to lack of network coverage, and the large amount of traffic generated by video transmission and the cost cannot be controlled.

[0028] The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0029] For ease of understanding, the specific process of the embodiment of the present invention is described below. Figure 1 , the first embodiment of the artificial intelligence data collection method in the embodiment of the present invention includes: 101. Deploy multiple edge device terminals in the target field scene, and configure and set up the multiple device terminals; In this embodiment, before using the acquisition module of the edge device terminal deployed in the target outdoor scene to collect the frame image in the target outdoor scene, the deployment quantity and deployment structure of the edge device terminal are determined according to the size of the target outdoor scene; Determining the number and structure of edge device terminals to be deployed according to the size of the target outdoor scene includes: deploying edge device terminals in a multi-node one-center topology structure in the target outdoor scene to ensure that monitoring can cover all scenes; The multiple edge device terminals after deployment are set up for collection; wherein the multiple edge device terminals after collection setting are used to collect frame images in the target outdoor scene.

[0030] The collection setting of the multiple edge device terminals after deployment includes: Set the corresponding acquisition parameters for each of the multiple edge device terminals after deployment, and configure the corresponding image push addresses for the multiple edge device terminals; wherein the acquisition parameters include at least one of the acquisition time period and the acquisition interval frame rate; the image push address is used to transmit the frame image to the corresponding acquisition platform.

[0031] 102. Using a collection module in an edge device terminal to collect frame images in a target outdoor scene; Based on the collection time period and collection interval frame rate set by the edge device terminal, the frame images of the target outdoor scene are collected using the camera or other sensor devices in the edge device terminal.

[0032] 103. Preprocess the collected frame images in the target outdoor scene using a processing module in the edge device terminal to obtain a preprocessed frame image; In this embodiment, the processing module in the edge device terminal performs filtering processing on the collected frame images to meet preset requirements to obtain filtered frame images, thereby reducing data redundancy.

[0033] Wherein, the processing module in the edge device terminal performs filtering processing on the collected frame images to meet preset requirements, including: at least one of deduplication processing and filtering processing for obtaining target images; The processing module in the edge device terminal performs deduplication processing on the collected frame images, including: extracting feature vectors of the collected continuous frame images using a convolutional neural network, calculating the similarity of the continuous frame images based on the feature vectors of the continuous frame images, and performing deduplication processing when the similarity is higher than a preset value to reduce data redundancy; in, Indicates the nth frame of the picture; represents the n+1th frame of the picture; represents the dot product of the vector; The processing module in the edge device terminal performs filtering processing on the current frame image to obtain the target image, including: Build a target detection model, use the target detection model to identify the target in the frame image, filter the image frames where the target is not identified, and remove the frame images that do not meet the requirements, thereby achieving the effect of reducing data redundancy.

[0034] 104. Using the transmission module in the edge device terminal to transmit the pre-processed frame image to the corresponding acquisition platform; In this embodiment, first, the edge device terminal establishes a communication connection with the acquisition platform to ensure the reliability and security of data transmission; then the transmission module in the edge device terminal transmits the preprocessed frame image to the corresponding acquisition platform according to the image push address.

[0035] 105. The acquisition platform uses the receiving and storage module to receive and save the pre-processed frame images, so that the sample engineer can uniformly screen the data.

[0036] This embodiment uses the above technical means to achieve low-cost and high-efficiency artificial intelligence data collection in field scenarios, reduce project costs, save traffic, reduce the workload of sample engineers, and improve work efficiency.

[0037] See also Figure 2 , the second embodiment of the artificial intelligence data collection method in the embodiment of the present invention includes: 201. Deploy edge devices with 4G / 5G in the target field scene and start the edge devices with 4G / 5G; In this embodiment, an edge device with 4G / 5G is deployed every 100 meters to ensure that monitoring can cover all scenarios.

[0038] 202. Set the edge device parameters with 4G / 5G, including: collection time period, collection interval frame rate; and configure the corresponding image push address; In this embodiment, the collection time period is set to 8 am to 5 pm every day, the collection interval frame rate is one frame every 5 seconds, and the image push address is set to the address of the central server to prepare for the real-time push of the collected data to the central server; 203. Enable the deduplication function of edge devices with 4G / 5G to reduce the transmission of redundant data; In this embodiment, two deduplication requirements are set according to user needs, including strict deduplication requirements and loose deduplication requirements; when strict deduplication requirements are selected, the features of the two pictures before and after are compared with the ones with higher similarity for filtering; when loose deduplication requirements are selected, the features of the two pictures before and after are compared with the ones with lower similarity for filtering; different deduplication modes are selected according to the corresponding scenarios.

[0039] 204. Determine whether to enable a filtering function for acquiring a target image based on whether the user performs targeted collection; In this embodiment, when the user performs targeted collection, the collected picture frames need to be filtered and processed in a targeted manner to filter out unnecessary picture frames; for example, when detecting human figures in the wild, the human figure detection model can be used to identify them first, and the images without human figures can be filtered out, thereby saving data transmission traffic and saving time for sample engineering to screen samples; if there are no requirements for the collected samples, the general model algorithm can be used for collection. The general model does not detect specific objects, but collects frame pictures according to the set rules. It is used more in the early sample collection of the algorithm; 205. Start the collection task, and use the edge device with 4G / 5G to deduplicate the collected frame images and / or obtain the filtering process of the target images to reduce redundant data; 206. The edge device with 4G / 5G sends the collected frame images to the collection platform in real time, making it convenient for sample engineers to uniformly screen the data.

[0040] This embodiment can be applied in fields including outdoor security, and the surrounding area can be controlled through an artificial intelligence data collection method provided by the present invention.

[0041] This embodiment successfully solves the problem of high cost and low efficiency in field scenarios by using 4G / 5G for communication and only collecting frame images. Through algorithm filtering and automatic filtering, the amount of data transmission and the collection of redundant data are reduced, which reduces project costs and traffic consumption, and also reduces the workload of sample engineers and improves work efficiency.

[0042] See also Figure 3 , a third embodiment of the artificial intelligence data collection method in the embodiment of the present invention includes: 301. Deploy multiple edge device terminals in the target farmland to ensure that monitoring can cover all target farmland; In this embodiment, the edge device terminal is configured with an edge algorithm box; the edge algorithm box is a processing module deployed on the edge device terminal, and is used to perform real-time processing and analysis on the collected data.

[0043] Among them, the edge device terminal collects the target farmland environment picture, and transmits the collected target farmland environment picture to the edge algorithm box for processing.

[0044] 302. Set the edge device terminal parameters, including: collection time period and collection interval frame rate; and configure the corresponding image push address; 303. Enable the edge device terminal to remove duplicates and obtain the filtering function of the target image to reduce the transmission of redundant data; In this embodiment, deduplication processing can help identify and remove duplicate frame images, thereby reducing storage and transmission costs. The filtering processing of obtaining the target image can filter out frame images that meet the requirements according to specific conditions, thereby improving data quality and accuracy; The deduplication process includes: extracting feature vectors of the collected continuous frame images using a convolutional neural network, calculating the similarity of the continuous frame images based on the feature vectors of the continuous frame images, and performing deduplication process when the similarity is higher than a preset value; Image processing techniques can also be used to filter out unnecessary information or noise.

[0045] 304. Start the collection task, and use the edge algorithm box configured on the edge device terminal to remove duplicates from the collected frame images and obtain filtering processing of the target images; 305. The edge device terminal transmits the processed frame image to the acquisition platform according to the configured image push address; 306. Based on the frame images received by the acquisition platform, the conditions of the farmland are monitored in real time, including soil moisture, vegetation growth, etc.

[0046] In this embodiment, the collected images are preprocessed using image processing technology, including operations such as denoising and contrast enhancement; the characteristics of the soil area are extracted by methods such as color analysis or texture analysis; the soil moisture prediction model after training is used to predict the soil moisture of the preprocessed images; Use image processing technology to segment vegetation areas and extract features, such as vegetation coverage and vegetation index. Analyze vegetation growth conditions, such as vegetation density and health status, based on vegetation characteristics and color information. Combine historical data and environmental factors to establish a vegetation growth model and predict vegetation growth areas.

[0047] This embodiment can be applied in the fields of environmental monitoring and agricultural Internet of Things, and the changes and trends of the environment can be analyzed by an artificial intelligence data collection method provided by the present invention. In the field of agricultural Internet of Things, agricultural management can be carried out accurately to improve agricultural production efficiency and quality.

[0048] The artificial intelligence data collection method in the embodiment of the present invention is described above. The artificial intelligence data collection device in the embodiment of the present invention is described below. Figure 4 , an embodiment of the artificial intelligence data acquisition device in the embodiment of the present invention includes: The edge device terminal acquisition module 401 is used to collect frame images in the target outdoor scene; In this embodiment, frame images in the target outdoor scene are collected based on sampling parameters of the edge device terminal, including: a collection time period and a collection interval frame rate.

[0049] The edge device terminal processing module 402 is used to pre-process the collected frame images in the target outdoor scene to obtain the pre-processed frame images; In this embodiment, the edge device terminal processing module 402 performs deduplication processing on the collected frame images and obtains filtering processing of the target images, thereby reducing data redundancy.

[0050] The edge device terminal processing module 402 performs deduplication processing on the collected frame images, including: extracting feature vectors of the collected continuous frame images using a convolutional neural network, calculating the similarity of the continuous frame images based on the feature vectors of the continuous frame images, and performing deduplication processing when the similarity is higher than a preset value to reduce data redundancy; The edge device terminal processing module 402 performs filtering processing on the deduplicated frame images to obtain the target images, including: Build a target detection model, use the target detection model to identify the target in the frame image, filter the image frames where the target is not identified, and remove the frame images that do not meet the requirements, thereby achieving the effect of reducing data redundancy.

[0051] The edge device terminal transmission module 403 transmits the pre-processed frame image to the corresponding acquisition platform; In this embodiment, first, the edge device terminal establishes a communication connection with the acquisition platform to ensure the reliability and security of data transmission; then the transmission module 403 in the edge device terminal transmits the preprocessed frame image to the corresponding acquisition platform based on the image push address.

[0052] The acquisition platform receiving and storage module 404 is used to receive and save the pre-processed frame images to facilitate the sample engineer to uniformly screen the data.

[0053] In this embodiment, the amount of data transmission is greatly reduced by collecting frame images; some unnecessary content is filtered out according to its own needs, thereby improving the accuracy of the sample; by automatically collecting frame images and sending them to the collection platform in real time, the workload of sample engineers is greatly reduced and work efficiency is improved.

[0054] See also Figure 5 Another embodiment of the artificial intelligence data acquisition device in the embodiment of the present invention includes: The edge device terminal acquisition module 501 is used to collect frame images in the target outdoor scene; The edge device terminal processing module 502 is used to pre-process the collected frame images in the target outdoor scene to obtain the pre-processed frame images; The edge device terminal transmission module 503 is used to transmit the pre-processed frame image to the corresponding acquisition platform; The acquisition platform receiving and storage module 504 is used to receive and save the pre-processed frame images; The acquisition platform analysis module 505 is used to monitor the conditions of the farmland in real time based on the frame images received by the acquisition platform, including soil moisture, vegetation growth, etc.

[0055] In this embodiment, the edge device terminal acquisition module 501 is used to collect target farmland environment pictures; In this embodiment, the edge device terminal processing module 502 includes: A deduplication unit 5021 is used to extract feature vectors of the collected continuous frame images using a convolutional neural network, calculate the similarity of the continuous frame images based on the feature vectors of the continuous frame images, and perform deduplication processing when the similarity is higher than a preset value, thereby reducing storage and transmission costs; The filtering unit 5022 is used to filter out frames that meet the requirements according to specific conditions to improve data quality and accuracy; In this embodiment, the acquisition platform analysis module 505 includes: The soil moisture monitoring unit 5051 is used to pre-process the collected images using image processing technology, including operations such as denoising and contrast enhancement; extract the characteristics of the soil area by methods such as color analysis or texture analysis; and predict the soil moisture of the pre-processed images using the trained soil moisture prediction model; The vegetation growth monitoring unit 5052 is used to segment and extract features of vegetation areas using image processing technology, such as vegetation coverage, vegetation index, etc.; analyze vegetation growth conditions such as vegetation density, health status, etc. based on vegetation characteristics and color information; and establish a vegetation growth model based on historical data and environmental factors to predict vegetation growth areas.

[0056] above Figure 4 and Figure 5 The artificial intelligence data acquisition device in the embodiment of the present invention is described in detail from the perspective of modular functional entities, and the electronic device in the embodiment of the present invention is described in detail from the perspective of hardware processing.

[0057] Figure 6 6 is a schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. The electronic device 600 may have relatively large differences due to different configurations or performances, and may include one or more processors (central processing units, CPU) 610 (for example, one or more processors) and a memory 620, and one or more storage media 630 (for example, one or more mass storage devices) storing application programs 633 or data 632. Among them, the memory 620 and the storage medium 630 can be short-term storage or permanent storage. The program stored in the storage medium 630 may include one or more modules (not shown in the figure), and each module may include a series of instruction operations in the electronic device 600. Furthermore, the processor 610 may be configured to communicate with the storage medium 630 to execute a series of instruction operations in the storage medium 630 on the electronic device 600.

[0058] The electronic device 600 may also include one or more power supplies 640, one or more wired or wireless network interfaces 650, one or more input and output interfaces 650, and / or one or more operating systems 631, such as Windows Serve, Mac OS X, Unix, Linux, FreeBSD, etc. It will be appreciated by those skilled in the art that Figure 6 The structure of the electronic device shown does not constitute a limitation on the electronic device, and may include more or less components than shown in the figure, or combine some components, or arrange the components differently.

[0059] The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. Instructions are stored in the computer-readable storage medium. When the instructions are executed on a computer, the computer executes the steps of the artificial intelligence data collection method.

[0060] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the above-described system, device, or unit can refer to the corresponding process in the aforementioned method embodiment and will not be repeated here.

[0061] 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 invention is essentially or the part that contributes to the prior art or the whole or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. 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.

[0062] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features thereof may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An artificial intelligence data collection method, characterized in that: include: Utilize the acquisition module of the edge device terminal deployed in the target outdoor scene to collect frame images in the target outdoor scene; Using the processing module in the edge device terminal to preprocess the collected frame images in the target outdoor scene to obtain the preprocessed frame images; The preprocessed frame image is transmitted to the corresponding acquisition platform by using the transmission module in the edge device terminal, so that the acquisition platform can analyze and process the preprocessed frame image to obtain the data acquisition result.

2. The artificial intelligence data collection method according to claim 1, characterized in that: Before using the acquisition module of the edge device terminal deployed in the target outdoor scene to acquire the frame image in the target outdoor scene, the method further includes: Provide multiple edge devices; Set a corresponding edge device terminal topology structure according to the target field scene, and deploy multiple edge device terminals in the target field scene according to the set topology structure; The multiple edge device terminals after deployment are set up for collection; wherein the multiple edge device terminals after collection setting are used to collect frame images in the target outdoor scene.

3. The artificial intelligence data collection method according to claim 2, characterized in that: The collection setting of the multiple edge device terminals after deployment includes: Setting the acquisition parameters corresponding to each of the multiple edge device terminals after deployment, and configuring the corresponding image push addresses for the multiple edge device terminals; wherein the acquisition parameters include at least one of the acquisition time period and the acquisition interval frame rate; the image push address is used to transmit the frame image to the corresponding acquisition platform; Accordingly, the collecting of frame images in the target outdoor scene by using a collection module of an edge device terminal deployed in the target outdoor scene includes: Based on the acquisition parameters corresponding to multiple edge device terminals, frame images in the target outdoor scene are collected; Accordingly, the method of using the transmission module in the edge device terminal to transmit the pre-processed frame image to the corresponding acquisition platform includes: Utilize the transmission module in the edge device terminal to transmit the preprocessed frame image to the corresponding acquisition platform based on the image push address.

4. The artificial intelligence data collection method according to claim 1, characterized in that: The method of preprocessing the collected frame images in the target outdoor scene using the processing module in the edge device terminal to obtain the preprocessed frame images includes: filtering the collected frame images in the target outdoor scene to meet preset requirements using the processing module in the edge device terminal to obtain the filtered frame images; The collected frame images in the target outdoor scene are filtered to meet preset requirements, including at least one of deduplication processing and filtering processing to obtain the target image.

5. The artificial intelligence data collection method according to claim 4, characterized in that: The method of using a processing module in an edge device terminal to perform deduplication processing on the collected frame images in the target outdoor scene includes: The convolutional neural network is used to extract the feature vectors of the collected continuous frame images, and the similarity of the continuous frame images is calculated based on the feature vectors of the continuous frame images. When the similarity is higher than the preset value, deduplication processing is performed; in, Indicates the nth frame of the picture; represents the n+1th frame of the picture; represents the dot product of the vector; Represents the magnitude of a vector.

6. The artificial intelligence data collection method according to claim 1, characterized in that: The method of using the transmission module in the edge device terminal to transmit the preprocessed frame image to the corresponding acquisition platform so that the acquisition platform analyzes and processes the preprocessed frame image to obtain data acquisition results includes: The transmission module in the edge device terminal is used to transmit the preprocessed frame image to the corresponding acquisition platform, so that the analysis module in the acquisition platform can extract features of the preprocessed frame image, generate data acquisition results based on the corresponding extracted features, and enable the acquisition platform to display the data acquisition results.

7. An artificial intelligence data collection device, characterized in that: include: The edge device terminal acquisition module is used to collect frame images in the target field scene; The edge device terminal processing module is used to pre-process the collected frame images in the target outdoor scene to obtain the pre-processed frame images; The edge device terminal transmission module is used to transmit the pre-processed frame images to the corresponding acquisition platform; The acquisition platform receiving and storage module is used to receive and save the pre-processed frame images.

8. The artificial intelligence data acquisition device according to claim 7, characterized in that: Set a corresponding edge device terminal topology structure according to the target field scene, and deploy multiple edge device terminals in the target field scene with the set topology structure; Configure and set up multiple edge device terminals, including: Set the terminal parameters of multiple edge devices, including: collection time period and collection interval frame rate; At the same time, configure corresponding data push platforms for multiple edge device terminals.

9. An electronic device, comprising a memory and at least one processor, characterized in that: Instructions are stored in the memory; The at least one processor calls the instructions in the memory so that the electronic device executes each step of the artificial intelligence data collection method as described in any one of claims 1 to 6.

10. A computer-readable storage medium having instructions stored thereon, characterized in that: When the instructions are executed by the processor, the various steps of the artificial intelligence data collection method as described in any one of claims 1 to 6 are implemented.

Images