Method, system and device for improving screenshot transmission and storage efficiency in cloud device and storage medium
By extracting screenshot feature vectors from cloud devices and comparing their similarity with cached images, and storing the difference image data, the problem of wasted storage space and network latency in storing massive screenshots is solved, achieving high-efficiency storage and transmission.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- CHUANGKE ZHILIAN (SHENZHEN) ELECTRONIC INFORMATION CO LTD
- Filing Date
- 2026-04-08
- Publication Date
- 2026-07-03
AI Technical Summary
In cloud devices, storing massive amounts of screenshots presents problems such as low storage space utilization, high metadata management overhead, and significant network latency and protocol overhead. This is especially true in scenarios such as cross-border e-commerce, where screenshots are used as data evidence, resulting in huge metadata management overhead and leading to performance bottlenecks in the storage system.
By obtaining the feature vector of the screenshot, a similarity comparison is made with the feature vector of the pre-cached image. For images with high similarity, only the difference image data is stored, while images with low similarity are compressed and stored. The difference image data is extracted using a pre-built large model and stored on the backend device.
It effectively reduces storage space usage, improves storage and transmission efficiency, saves storage costs, and reduces network requests and computing resource consumption.
Smart Images
Figure CN122340221A_ABST
Abstract
Description
[0001] Technology Neighborhood This invention relates to the field of cloud device storage technology, and in particular to methods, systems, devices and storage media for improving the efficiency of screenshot transmission and storage in cloud devices. Background Technology
[0002] In scenarios such as mobile office, online education, and cloud gaming, massive screenshot storage for cloud phones / cloud computers is a necessity, especially in cross-border e-commerce. Screenshots serve as data evidence for monitoring and archiving. When the daily screenshot volume of cloud phones / cloud computers exceeds ten million, metadata consumes up to 120GB of memory, with retrieval latency exceeding 5 seconds (test data from a traditional EXT4 file system). Massive image storage places extremely high demands on the storage system, resulting in huge metadata management overhead. Each file (regardless of size) requires a corresponding metadata structure (such as an inode) in the file system to store its attributes (filename, size, permissions, timestamp, physical block pointer, etc.). Massive numbers of small files mean managing a massive number of metadata entries. Storage space utilization decreases, as the file system allocates storage space in fixed-size blocks or clusters. A 1KB file in a 4KB block system will occupy the entire 4KB block, resulting in 3KB of wasted space (internal fragmentation). Massive numbers of small files accumulate to a huge amount of wasted space. Distributed systems amplify overhead. In distributed file systems (such as HDFS) or object storage, accessing each small file typically requires the client to communicate with a metadata server to obtain location information, and then communicate with a data server to retrieve the data. A massive number of small files translates to a massive number of network requests, making the metadata server a bottleneck, and significantly amplifying network latency and protocol overhead. Summary of the Invention
[0003] The purpose of this invention is to address the technical problems existing in the background art by proposing a method, system, device, and storage medium for improving the efficiency of screenshot transmission and storage in cloud devices.
[0004] To achieve the above-mentioned technical objectives, the technical solution adopted by the present invention is as follows: A first implementation of the first aspect of the present invention provides a method for improving the efficiency of screenshot transmission and storage in cloud devices, comprising: S101. Obtain a screenshot of the cloud device and send the screenshot to a pre-set large model for processing to extract the screenshot feature vector; S102. Compare the similarity between the screenshot feature vector and all pre-cached image feature vectors, wherein each image feature vector is extracted from multiple pre-cached source images. S103. Process the high-similarity screenshot feature vector with the corresponding image feature vector to extract the difference image data, and store the difference image data in a preset backend storage device, wherein high similarity is higher than or equal to a preset threshold.
[0005] Optionally, in a second implementation of the first aspect of the present invention, the screenshot feature vector is a low-dimensional, dense, and semantically rich numerical vector that is transformed from the original pixel matrix of the screenshot.
[0006] Optionally, in a third implementation of the first aspect of the present invention, step S103 includes: S1031. Determine whether the similarity is less than a preset threshold; S1032. If so, cache the current screenshot feature vector, compress the corresponding screenshot and store it in the preset backend storage device; S1033. If not, extract the difference image data between the current screenshot feature vector and the corresponding image feature vector, record the index data of the difference image data and the corresponding source image, and store the difference image data in the backend storage device.
[0007] Optionally, in a fourth implementation of the first aspect of the present invention, the method for improving the efficiency of screenshot transmission and storage in the cloud device further includes: S104. Perform derivation processing on the difference image data and the corresponding source image to obtain the derived image.
[0008] Optionally, in a fifth implementation of the first aspect of the present invention, the format of the differential image data includes a metadata area and a differential image data area. The metadata area is used to record index data of the source image, the name, attributes, size and creation time of the derived image.
[0009] A second aspect of the present invention provides a system for improving the efficiency of screenshot transmission and storage in cloud devices, comprising: The acquisition module is used to acquire screenshots of cloud devices and send the screenshots to a pre-set large model for processing to extract the screenshot feature vectors; The comparison module is used to compare the similarity between the screenshot feature vector and all pre-cached image feature vectors, where each image feature vector is extracted from multiple pre-cached source images; The storage module is used to process the feature vectors of highly similar screenshots with the corresponding image feature vectors to extract the difference image data and store the difference image data in a preset backend storage device, wherein high similarity is greater than or equal to a preset threshold.
[0010] A first implementation of the third aspect of the present invention provides a device for improving the efficiency of screenshot transmission and storage in a cloud device. The device for improving the efficiency of screenshot transmission and storage in a cloud device includes: a memory and at least one processor. The memory stores instructions, and the memory and the at least one processor are interconnected via a line. The at least one processor invokes the instructions in the memory to cause the screenshot transmission and storage efficiency improvement device in the cloud device to execute the screenshot transmission and storage efficiency improvement method in the cloud device as described in any one of the first aspects of the present invention.
[0011] A first implementation of the fourth aspect of the present invention provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the method for improving screenshot transmission and storage efficiency in a cloud device as described in any one of the first aspects of the present invention.
[0012] Compared with the prior art, the present invention has the following beneficial technical effects: by comparing the similarity between the screenshot feature vector and the feature vector of all pre-cached images, and extracting the corresponding difference image data, the difference image data is stored in a pre-set backend storage device, thereby reducing the storage space occupied, improving storage efficiency and transmission efficiency, and by storing the source image and difference image data instead of the existing solution of storing all images, storage costs are saved for cloud devices. Attached Figure Description
[0013] Figure 1 This is a schematic diagram of the first embodiment of the method for improving the efficiency of screenshot transmission and storage in cloud devices according to the present invention; Figure 2 This is a schematic diagram of the third embodiment of the method for improving the efficiency of screenshot transmission and storage in cloud devices according to the present invention; Figure 3 This is a schematic diagram of the fourth embodiment of the method for improving the efficiency of screenshot transmission and storage in cloud devices according to the present invention; Figure 4 This is a schematic diagram illustrating the formation of derived images in the method for improving the efficiency of screenshot transmission and storage in cloud devices according to an embodiment of the present invention; Figure 5 This is a schematic diagram of the structure of differential image data in the method for improving the efficiency of screenshot transmission and storage in cloud devices according to an embodiment of the present invention; Figure 6 A schematic diagram of an embodiment of the screenshot transmission and storage efficiency improvement system in a cloud device according to an embodiment of the present invention; Figure 7 This is a schematic diagram of an embodiment of a device for improving the efficiency of screenshot transmission and storage in a cloud device according to an embodiment of the present invention. Detailed Implementation
[0014] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other.
[0015] For ease of understanding, the specific process of the embodiments of the present invention is described below. Please refer to [link / reference]. Figure 1 - Appendix Figure 5 The method for improving screenshot transmission and storage efficiency in cloud devices according to embodiments of the present invention includes: S101. Obtain a screenshot of the cloud device and send the screenshot to a pre-set large model for processing to extract the screenshot feature vector; In this embodiment, cloud devices refer to virtual machines, container instances, or cloud-hosted IoT devices. Screenshots of the current screen are captured from the cloud devices using tools such as APIs or remote control software. The pre-set large model is a pre-trained AI large model that has the function of extracting feature vectors from images. Therefore, the captured screenshot is sent to the corresponding AI large model, and the corresponding screenshot feature vector is obtained.
[0016] Furthermore, the screenshot feature vector transforms the original pixel matrix of the screenshot into a low-dimensional, dense, and semantically rich numerical vector.
[0017] S102. Compare the similarity between the screenshot feature vector and all pre-cached image feature vectors, wherein each image feature vector is extracted from multiple pre-cached source images. In this embodiment, the system has a pre-arranged storage space for storing various source images. There are two scenarios: First, the system is in its initial state, meaning there are no screenshots in the storage space. When a user takes a screenshot for the first time using the cloud device, that screenshot becomes the first batch of source images in the storage space. Subsequent screenshots by the user can be compared with the source images already stored in the storage space for similarity. Second, multiple source images are pre-stored in the storage space. These source images can be screenshots taken from the user's own mobile phone, computer, or other local electronic devices. Subsequent screenshots are then compared with the source images for similarity. Among them, similarity comparison calculation can be performed using cosine similarity calculation, Euclidean distance, Manhattan distance, etc.
[0018] S103. Process the high-similarity screenshot feature vector with the corresponding image feature vector to extract the difference image data, and store the difference image data in a preset backend storage device, wherein high similarity is higher than or equal to a preset threshold. In this embodiment, the preset threshold is actually the image similarity threshold, which can be set to 80%, 90%, etc. According to the user's needs, if higher accuracy is required, the image similarity threshold can be set to a higher threshold, such as 90%, 95%, etc. By comparing the extracted new image feature vector (i.e., screenshot feature vector) with the feature vector of the cached old image (i.e., image feature vector), if the image similarity is greater than or equal to the set threshold, the difference image data of the new image is extracted through feature vector and new and old image processing, and the index of the old image compared with it is recorded.
[0019] Furthermore, step S103 specifically includes: S1031. Determine whether the similarity is less than a preset threshold; S1032. If so, cache the current screenshot feature vector, compress the corresponding screenshot and store it in the preset backend storage device; S1033. If not, extract the difference image data between the current screenshot feature vector and the corresponding image feature vector, record the index data of the difference image data and the corresponding source image, and store the difference image data in the backend storage device.
[0020] In this embodiment, a similarity less than a preset threshold is considered low similarity, and a similarity greater than or equal to the preset threshold is considered high similarity. The threshold setting is determined by the user requirements or system resource configuration requirements mentioned above. After similarity comparison, the screenshot feature vector with low similarity is treated as a new image. Therefore, the screenshot corresponding to the screenshot feature vector is directly compressed to reduce storage space occupation and computing resources. The compressed screenshot is stored in the backend storage device as a member of the old image and will be used as the object for similarity comparison of subsequent new screenshots. The screenshot feature vector with high similarity is treated as a derivative of the old image. If such high similarity screenshots are stored in the backend storage device at the same size, it will occupy a lot of storage resources. Therefore, it is necessary to differentiate the high similarity screenshot feature vector with the corresponding image feature vector to find the difference points. Pixel-level difference masks can be used to mark the difference areas, and then the corresponding difference image data can be extracted. It should be noted that this difference image data needs to record the index data of the corresponding source image for subsequent retrieval work. This differentiated processing approach allows for various scenarios. For example, in high-similarity scenarios, the difference image data and old image indexes are stored to enable subsequent reconstruction of new images from old images and difference data, preventing data loss. In low-similarity scenarios, the complete new image and cached new image features / images are stored to facilitate rapid comparison of subsequent new images and reduce redundant storage. For highly similar images (such as multiple screenshots of the same interface or images with only minor modifications), only the difference data is stored, not the complete image. This significantly reduces storage resource consumption, especially in scenarios such as screenshots, monitoring screens, and version iteration images. Furthermore, caching old image feature vectors eliminates the need to read the complete image file during new image comparison; only high-dimensional vectors need to be compared (which is computationally faster), reducing I / O overhead, improving judgment efficiency, and ultimately enhancing storage and retrieval efficiency.
[0021] S104. Perform derivation processing on the difference image data and the corresponding source image to obtain the derived image.
[0022] In this embodiment, the backend storage device stores two types of data: original screenshot data and difference image data. The cloud device uploads the captured image or the extracted difference image data to the backend storage device. The size of the processed difference image data is much smaller than that of the original image (e.g., if the similarity reaches 90%, only 10% of the data of the original screenshot needs to be stored, occupying only 1 / 10 of the original storage space), and the storage space occupied is also smaller. The new image generated by the source image through the difference image data and the source image can also be called a derived image. Therefore, there will be multiple source images and a large amount of difference image data in a single backend storage device.
[0023] In practice, when a user initiates an image access request, if the accessed image is the source image of the storage system, the image access module returns the access URL of the source image to the user, who can then directly access the source image via the URL. If the accessed image is a derived image, the module uses the difference image data and the recorded original image index to generate a derived image from the difference image data and the source image, and returns the derived image URL to the user, who can then directly access the derived image via the URL.
[0024] Furthermore, the format of the differential image data includes a metadata area and a differential image data area. The metadata area is used to record the index data of the source image, the name, attributes, size, and creation time of the derived image.
[0025] In this embodiment, by extracting and transmitting the difference image data of the images, the problem of large network bandwidth being occupied by the transmission of a large number of images is solved, and the network transmission efficiency is improved. By storing the source image and the difference image data instead of the existing solution of storing all images, the problem of excessive storage space occupied by massive image storage is solved, and the storage efficiency is improved. In the scenario of massive cloud mobile phone or cloud computer image storage, the cost is saved by tens or even hundreds of times.
[0026] The above describes the method for improving screenshot transmission and storage efficiency in cloud devices according to embodiments of the present invention. The following describes the method for improving screenshot transmission and storage efficiency in cloud devices according to embodiments of the present invention. Please refer to [link / reference]. Figure 6 The screenshot transmission and storage efficiency improvement system in the cloud device includes: The acquisition module 201 is used to acquire screenshots of cloud devices and send the screenshots to a pre-set large model for processing to extract the screenshot feature vectors. The comparison module 202 is used to compare the similarity between the screenshot feature vector and all pre-cached image feature vectors, wherein each image feature vector is extracted from multiple pre-cached source images; Storage module 203 is used to process the high-similarity screenshot feature vector and the corresponding image feature vector to extract the difference image data and store the difference image data in a preset back-end storage device, wherein high similarity is higher than or equal to a preset threshold.
[0027] Among them, the screenshot feature vector is a low-dimensional, dense, and semantically rich numerical vector transformed from the original pixel matrix of the screenshot.
[0028] Specifically, storage module 203 can also perform the following: S1031. Determine whether the similarity is less than a preset threshold; S1032. If so, cache the current screenshot feature vector, compress the corresponding screenshot and store it in the preset backend storage device; S1033. If not, extract the difference image data between the current screenshot feature vector and the corresponding image feature vector, record the index data of the difference image data and the corresponding source image, and store the difference image data in the backend storage device.
[0029] The system for improving the efficiency of screenshot transmission and storage in cloud devices specifically includes: The derivation module 204 is used to perform derivation processing on the difference image data and the corresponding source image to obtain the derived image.
[0030] The format of the differential image data includes a metadata area and a differential image data area. The metadata area is used to record the index data of the source image, the name, attributes, size and creation time of the derived image.
[0031] In this embodiment, by extracting and transmitting the difference image data of the images, the problem of large network bandwidth being occupied by the transmission of a large number of images is solved, and the network transmission efficiency is improved. By storing the source image and the difference image data instead of the existing solution of storing all images, the problem of excessive storage space occupied by massive image storage is solved, and the storage efficiency is improved. In the scenario of massive cloud mobile phone or cloud computer image storage, the cost is saved by tens or even hundreds of times.
[0032] The method for improving screenshot transmission and storage efficiency in cloud devices in this embodiment of the invention is described in detail from the perspective of unitized functional entities. The device for improving screenshot transmission and storage efficiency in cloud devices in this embodiment of the invention is described in detail from the perspective of hardware processing.
[0033] Figure 7 This is a schematic diagram of a screenshot transmission and storage efficiency improvement device in a cloud device according to an embodiment of the present invention. This screenshot transmission and storage efficiency improvement device in the cloud device can vary significantly due to different configurations or performance. It may include one or more central processing units (CPUs) 310 (e.g., one or more processors) and memory 320, and one or more storage media 330 (e.g., one or more mass storage devices) for storing application programs 333 or data 332. The memory 320 and storage media 330 can be temporary or persistent storage. The program stored in the storage media 330 may include one or more units (not shown in the diagram), each unit may include a series of instruction operations on the screenshot transmission and storage efficiency improvement device in the cloud device. Furthermore, the processor 310 may be configured to communicate with the storage media 330 to execute a series of instruction operations on the storage media 330 on the screenshot transmission and storage efficiency improvement device in the cloud device.
[0034] The device for improving screenshot transmission and storage efficiency in cloud devices may also include one or more power supplies 340, one or more wired or wireless network interfaces 350, one or more input / output interfaces 360, and / or one or more operating systems 331, such as Windows Server, Mac OS X, Unix, Linux, FreeBSD, etc. Those skilled in the art will understand that... Figure 7 The structure of the cloud device shown, which improves the efficiency of screenshot transmission and storage, does not constitute a limitation on communication protocol devices based on local area network projection. It may include more or fewer components than shown, or combine certain components, or have different component arrangements.
[0035] The present invention also provides a computer-readable storage medium, which can be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, wherein the computer-readable storage medium stores instructions that, when the instructions are executed on a computer, cause the computer to perform the steps of the screenshot transmission and storage efficiency improvement method in the cloud device.
[0036] If the integrated unit is implemented as 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, in essence, 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. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.
[0037] The above describes a method or multiple implementations for improving screenshot transmission and storage efficiency in cloud devices, but it is not intended that the specific implementation of this invention is limited to these descriptions. Any methods or structures similar to or identical to those of this invention, or any technical deductions or substitutions made based on the concept of this invention, should be considered within the scope of protection of this invention.
Claims
1. A method for improving the efficiency of screenshot transmission and storage in cloud devices, characterized in that, include: S101. Obtain a screenshot of the cloud device and send the screenshot to a preset large model for processing to extract the screenshot feature vector; S102. Compare the similarity between the screenshot feature vector and all pre-cached image feature vectors, wherein each image feature vector is extracted from multiple pre-cached source images. S103. The screenshot feature vector with high similarity is processed with the corresponding image feature vector to extract the difference image data, and the difference image data is stored in a preset backend storage device, wherein the high similarity is higher than or equal to a preset threshold.
2. The method for improving screenshot transmission and storage efficiency in a cloud device according to claim 1, characterized in that, The screenshot feature vector is a low-dimensional, dense, and semantically rich numerical vector transformed from the original pixel matrix of the screenshot.
3. The method for improving screenshot transmission and storage efficiency in a cloud device according to claim 2, characterized in that, Step S103 includes: S1031. Determine whether the similarity is less than a preset threshold; S1032. If so, then cache the current screenshot feature vector, and compress and store the corresponding screenshot in a preset backend storage device; S1033. If not, extract the difference image data between the current screenshot feature vector and the corresponding image feature vector, record the difference image data and the index data of the corresponding source image, and store the difference image data in the backend storage device.
4. The method for improving screenshot transmission and storage efficiency in a cloud device according to claim 3, characterized in that, Also includes: S104. Perform derivative processing on the difference image data and the corresponding source image to obtain a derived image.
5. The method for improving the efficiency of screenshot transmission and storage in a cloud device according to claim 4, characterized in that, The format of the differential image data includes a metadata area and a differential image data area. The metadata area is used to record the index data of the source image, the name, attributes, size and creation time of the derived image.
6. A system for improving the efficiency of screenshot transmission and storage in cloud devices, characterized in that, include: The acquisition module is used to acquire screenshots of cloud devices and send the screenshots to a pre-set large model for processing to extract screenshot feature vectors; The comparison module is used to compare the similarity of the screenshot feature vector with all pre-cached image feature vectors, wherein each image feature vector is extracted from multiple pre-cached source images; The storage module is used to process the screenshot feature vector with high similarity with the corresponding image feature vector to extract the difference image data, and store the difference image data in a preset backend storage device, wherein the high similarity is higher than or equal to a preset threshold.
7. A device for improving the efficiency of screenshot transmission and storage in cloud devices, characterized in that, The cloud device for improving the efficiency of screenshot transmission and storage includes: a memory and at least one processor, wherein the memory stores instructions and the memory and the at least one processor are interconnected via a line; The at least one processor invokes the instructions in the memory to cause the screenshot transmission and storage efficiency improvement device in the cloud device to execute the screenshot transmission and storage efficiency improvement method in the cloud device as described in any one of claims 1-5.
8. A computer-readable storage medium storing a computer program thereon, characterized in that, When the computer program is executed by the processor, it implements the method for improving the efficiency of screenshot transmission and storage in cloud devices as described in any one of claims 1-5.