Safe closed-loop circulation and physical output control method and system for ultra-large raster image

Abstract

The invention discloses a secure closed-loop circulation and physical output control method and system for an oversized raster image. The method comprises the following steps: a client adaptively calculates fragmentation granularity according to a network state, fragments an image and encrypts and transmits the image; the server performs secondary encryption and distributed storage on the received data; the preview end constructs a multi-resolution hierarchy based on a viewport mapping technology, only extracts visual slices and superposes watermarks, and transmits the visual slices and the watermarks to a client memory for zero-cache rendering; and the printing end constructs a secure channel bound with a hardware fingerprint, carries out streaming decryption on the data, embeds traceability information, and injects a printing drive in real time. According to the method, the problems of unstable GB-level image transmission and unsmooth preview are solved, and high-efficiency circulation and strict protection of design assets are realized through full-process data non-landing streaming control.

Description

Technical Field

[0001] This invention relates to the fields of computer data processing and information security technology, and in particular to a secure closed-loop transfer and physical output control method and system for ultra-large raster image data. Specifically, this invention addresses efficient fragmented transmission, distributed encrypted storage, zero-buffered preview based on viewport mapping, and hardware-bound streaming physical output control technology for high-resolution bitmaps (such as inkjet prints and industrial design drawings) at the GB level and above in uncontrolled network environments. Background Technology

[0002] With the accelerated digital transformation of the advertising printing, industrial manufacturing, and architectural decoration industries, the precision of image data involved in the production process is constantly increasing. In practical applications, high-precision printing source files (such as TIFF, PSD, high-resolution PDF, etc.) often contain complex layers and color information, and single file sizes exceeding 1GB or even reaching 10GB have become commonplace. The flow and processing of such ultra-large raster image data poses a severe challenge to existing data management systems.

[0003] The search revealed that existing image processing and streaming technologies have the following main shortcomings:

[0004] 1. Low efficiency and poor stability in network transmission of extremely large files: Traditional common file transfer protocols (such as FTP and HTTP Multipart upload) typically employ whole-file streaming or fixed-size chunked transmission mechanisms. When dealing with gigabyte-scale files, network bandwidth fluctuations or momentary disconnections often lead to the failure of the entire transmission task, requiring retransmission from the beginning and severely wasting bandwidth resources and time. Furthermore, existing technologies lack adaptive adjustment mechanisms based on network conditions, failing to dynamically adjust packet segmentation strategies in weak network environments, resulting in extremely low transmission success rates.

[0005] 2. The preview process carries a serious risk of data leakage (“data persistence” issue): In existing workflows, if a very large image needs to be previewed or approved, the client typically needs to download the complete source file to its local disk first, and then use a local image viewer (such as Photoshop or Windows Photo Viewer) for rendering. This “download-and-preview” model has two major drawbacks: First, the download waiting time is too long, seriously affecting collaboration efficiency; second, once the source file is stored on the client's local disk (even in a temporary folder), it is no longer under the server's security control and is easily copied, tampered with, or removed by unauthorized personnel via removable storage devices, leading to the leakage of core intellectual property rights.

[0006] 3. Lack of closed-loop security control in the physical output stage: Current printing control processes are typically loosely coupled: documents are decrypted and restored to plaintext on the print server, then manually opened and sent to the printer by the operator. During the time window between "decryption" and "printing," the plaintext document is completely exposed to the operating system, posing a significant risk of interception. Furthermore, existing technology cannot effectively restrict document output to specific printers, cannot prevent internal personnel from copying confidential drawings to other unauthorized devices for printing, and the printing process lacks robust anti-counterfeiting and traceability measures (such as dynamic watermarks or encrypted QR codes) strongly linked to the document content.

[0007] In summary, existing technical solutions cannot simultaneously meet the requirements of efficient transmission, online zero-buffered preview, and secure closed-loop physical output for ultra-large raster image data. Therefore, a new technical solution is urgently needed that enables data to be transmitted without ever being stored on the ground, achieving high efficiency and controllability throughout the entire process. Summary of the Invention

[0008] To address the shortcomings of existing technologies, such as low transmission efficiency and susceptibility to interruption of gigabyte-sized raster image data in uncontrolled network environments, the need to download source files for previewing leading to data storage and potential leakage risks, and the lack of hardware-bound closed-loop security control during physical printing, this invention provides a secure closed-loop transfer and physical output control method and system for ultra-large raster image data. This invention aims to ensure that image data is never stored in plaintext form on the local disk at any stage of transmission, storage, preview, and printing through end-to-end data fragmentation, encrypted transfer, and streaming processing mechanisms, thereby effectively resolving the contradiction between large file transfer efficiency and core asset security.

[0009] To solve the above-mentioned technical problems, the technical solution provided by this invention is "a secure closed-loop transfer and physical output control method for ultra-large raster image data", which includes the following steps:

[0010] 1. Adaptive Access and Dynamic Slicing: The client obtains the ultra-large raster image file to be processed, monitors the bandwidth jitter rate and packet loss rate of the current network link in real time, and calculates the optimal dynamic slicing granularity based on the monitoring results; the image file is logically cut into multiple data slices, the hash feature value of each slice is calculated, and a metadata index containing the slice order and verification information is established. Then, the sliced ​​data is asynchronously transmitted to the server through an encrypted channel.

[0011] 2. Fragmentation and Reassembly with Encrypted Storage: The server receives fragmented data and verifies its integrity. After verification, it does not restore the original file but uses a distributed storage strategy to persistently store the fragmented data. At the same time, a global key uniquely bound to the image file is generated to perform secondary encryption on the stored fragmented data and generate a corresponding ciphertext index table to ensure the fragmentation and encryption of data at the static storage level.

[0012] 3. Zero-cache preview based on viewport mapping: In response to a preview request initiated by the client, the server constructs multi-resolution hierarchical data based on the image pyramid model, and calculates the set of image slices corresponding to the current visible area according to the viewport coordinates and zoom level fed back by the client; the server only extracts the encrypted slices corresponding to the set, dynamically decrypts and synthesizes digital watermarks in real time before transmitting them to the client; the client receives and renders the slice data in volatile memory, and releases the memory immediately after the preview ends or the viewport is switched, ensuring that the source file is not written to the disk.

[0013] 4. Streaming decryption and hardware-bound output: In response to physical printing commands, the server locks the hardware fingerprint information of the target physical printing device to build an end-to-end secure output channel; the server reads the stored encrypted fragments, performs streaming decryption and splicing of the image data according to the buffer size required by the print driver, dynamically embeds the production traceability QR code during the rasterization process, injects the processed data stream into the printer buffer in real time for physical output, and destroys the decrypted data immediately after output is completed.

[0014] Furthermore, the present invention also provides a secure closed-loop transfer and physical output control system for ultra-large raster image data that implements the above method, the system comprising:

[0015] 1. Client Interaction Module: This module provides a user interface for selecting files and performing dynamic file segmentation and uploading tasks based on network conditions. It includes a memory-based rendering engine that receives image slice streams from the server and displays them without buffering.

[0016] 2. Server-side core processing module: Used to receive fragmented data, perform hash verification, distributed storage and image pyramid generation, and is responsible for persistent data management and basic processing.

[0017] 3. Security Management Module: Used to manage global keys and permission policies, responsible for generating one-time session keys bound to the printer hardware fingerprint during the physical printing stage, and monitoring the decryption and destruction process of the data stream.

[0018] 4. Physical Output Driver Module: This module receives streaming decrypted data from the server, converts it into a print instruction set (such as PJL / PCL), and drives the physical printing engine to complete image output.

[0019] Compared with the prior art, the present invention has the following advantages:

[0020] 1. Significantly improves the stability and efficiency of large file transmission: By introducing a network state-aware adaptive fragmentation mechanism, this invention can dynamically adjust the transmission strategy (such as adjusting the fragment size and concurrency) in weak network environments. Combined with hash-based instant transmission technology, it completely solves the industry problem of frequent crashes during GB-level file transmission, and greatly improves the transmission success rate and bandwidth utilization.

[0021] 2. Achieved "zero trust" and "zero caching" in the preview process: Utilizing viewport mapping and image pyramid technology, the client only loads the visible area slices on demand and renders only in volatile memory, eliminating the risk of data leakage caused by downloading the complete source file to the local disk in the traditional mode, and realizing a secure preview experience of "read and burn".

[0022] 3. A robust security loop for physical output has been established: Through streaming decryption injection and hardware fingerprint binding technology, it is ensured that the file is only restored the moment it reaches the memory of a specific printer, preventing man-in-the-middle attacks from intercepting plaintext files or illegally transferring files to other devices for printing; at the same time, with the help of dynamic traceability QR codes, the entire production process is managed, controllable, and traceable, effectively protecting the company's core intellectual property rights.

[0023] In summary, existing technical solutions cannot simultaneously meet the requirements of efficient transmission, online zero-buffered preview, and secure closed-loop physical output for ultra-large raster image data. Therefore, a new technical solution is urgently needed that enables data to be transmitted without ever being stored on the ground, achieving high efficiency and controllability throughout the entire process. Attached Figure Description

[0024] Figure 1 This is a diagram of the overall system architecture in an embodiment of the present invention;

[0025] Figure 2 This is an overall business process diagram in an embodiment of the present invention;

[0026] Figure 3 This is a schematic diagram of the adaptive dynamic sharding logic in an embodiment of the present invention;

[0027] Figure 4 This is a schematic diagram of the zero-buffered preview principle based on viewport mapping in an embodiment of the present invention;

[0028] Figure 5 This is a timing diagram of hardware binding and streaming decryption in an embodiment of the present invention. Detailed Implementation

[0029] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments. It should be noted that these embodiments are only for explaining the present invention and do not constitute a limitation on the scope of protection of the present invention.

[0030] 1. System Architecture and Hardware Environment

[0031] like Figure 1 As shown in the figure, the secure closed-loop transfer and physical output control system for ultra-large raster image data provided in this embodiment mainly consists of three parts: a client interaction subsystem, a server core processing subsystem, and a physical output control subsystem.

[0032] (1) Client Interaction Subsystem: Runs on PC terminals (such as Windows / macOS) and is encapsulated based on the Electron framework. It integrates a network status monitoring unit for real-time collection of the current link's RTT (Round Trip Time) and packet loss rate; an adaptive fragmentation calculation module for segmenting large files according to network status; and a core memory rendering engine, which is configured to allocate buffers only in volatile memory for receiving and displaying image slices, and prohibits writing any image data to the local hard drive or temporary folder.

[0033] (2) Server-side core processing subsystem: Deployed on a Linux server cluster, based on the Java Spring Boot architecture. It includes a distributed storage unit (for storing encrypted fragments), an image pyramid generation engine (for preprocessing large images to generate multi-resolution layers), and a security management module (responsible for key management and authorization verification).

[0034] (3) Physical Output Control Subsystem: This subsystem includes the print server and driver connected to the physical printer. It is configured to receive only streaming encrypted data from the server and has hardware fingerprint feedback capability.

[0035] 2. Adaptive Access and Slicing Process for Ultra-Large Image Data

[0036] like Figure 2 and Figure 3 As shown, when processing the upload of ultra-large raster images (such as TIFF or high-definition PDF) exceeding 1GB, this invention no longer uses the traditional whole-image upload method, but instead performs the following steps:

[0037] (1) Environmental Awareness: Before the client starts the upload task, the network status monitoring unit sends a probe packet to the server. Assume that the current bandwidth is measured to be B = 50Mbps and the packet loss rate is P = 0.5%.

[0038] (2) Dynamic fragmentation calculation: The system presets the basic fragmentation size S. base =10MB. According to the formula S... dynamic= S base Calculate the current optimal partition size S using the formula: ×(1-β·P) (where β is the penalty coefficient). dynamic Based on the calculation results, the transmission modes are divided into, for example: Figure 3 The three strategies shown: when S dynamic Approaching S base When the value fluctuates, strategy A (large granularity) is applied; when the value is below the preset threshold, strategy B is applied; when the value is below the preset threshold, strategy C (small granularity + high redundancy) is applied. If network fluctuations cause the packet loss rate to rise to 5%, the system will automatically reduce the size of subsequent fragments to 2MB to reduce retransmission overhead.

[0039] (3) Hash Indexing and Encryption: The client presses S dynamic The file is logically split, and the SHA-256 hash value of each fragment is calculated. Then, the fragments are encrypted using the session key using AES-256, generating an encrypted data packet {Enc(Shard)}. i ), Hash i Index i}

[0040] (4) Asynchronous Transmission and Instant Transmission: The client prioritizes sending a fragmented hash list. The server compares this list with the database; if the hash already exists, it returns a "successful instant transmission" flag. Otherwise, the client initiates multi-threaded asynchronous upload of the encrypted data packet. Upon receiving the packet, the server directly stores it in encrypted form in a distributed file system (such as MinIO) and updates the metadata index.

[0041] 3. Zero-buffered preview workflow based on viewport mapping

[0042] like Figure 4 As shown, in order to achieve secure preview without downloading the full image, this embodiment performs the following steps:

[0043] (1) Image Pyramid Construction: The server preprocesses the uploaded ultra-large image in the background to generate an N-level pyramid structure. Level 0 is the original image, and Level N is the smallest thumbnail. Each layer of the image is meshed and cut into 256×256 pixel slices.

[0044] (2) Viewport Request: Assume that the user is viewing a local detail of an image on the client, the current zoom level is L2, and the center coordinates of the screen's visible area (Viewport) are (x, y). The client calculates that the viewport covers the slices in rows (R2, C3) to (R5, C6) of the server's Level 2 hierarchy, and requests only the IDs of these 12 slices from the server.

[0045] (3) Dynamic synthesis and transmission: The server reads the corresponding encrypted slice, decrypts it, and dynamically overlays the "employee number + time" digital watermark of the current previewer in memory. After re-encryption, it is pushed to the client through Web Socket stream.

[0046] (4) Self-destructing after viewing: The client-side memory rendering engine receives the data stream and draws the image directly in the video memory. When the user drags the scroll bar or closes the window, the system immediately calls the memory overwrite instruction (such as memset) to forcibly clear the old slice data, ensuring that there are no remnants on the local disk at any time.

[0047] 4. Hardware-bound streaming decryption and printing process

[0048] like Figure 5 As shown, when a file needs to be physically output, the system first makes an intelligent judgment based on the user's operation path: if the user skips the preview step and initiates printing directly, the system will by default call the cloud thumbnail for simple confirmation, or directly enter the print authentication process; after confirmation, the following strict closed-loop control steps are executed:

[0049] (1) Hardware fingerprint locking: The client initiates a print command, and the server connects to the target printer to obtain its unique hardware fingerprint (in this embodiment, it is specifically reflected as the printer's factory serial number Device ID or the hash value of the network card MAC address to ensure the uniqueness and immutability of the device).

[0050] (2) Construct a secure channel: The server combines the file ID, the current timestamp, and the Device ID to generate a one-time print session key K. session .

[0051] (3) Streaming processing and source injection:

[0052] The server reads the encrypted fragment sequence from the storage area;

[0053] In the server's memory, the data is decrypted in blocks, and an image processing library (such as OpenCV / ImageMagick) is called to embed a high-strength encrypted QR code (containing the person who printed it, the time, and the document fingerprint) into the image rasterization data.

[0054] The process uses a "producer-consumer" model. After processing a chunk, it is immediately pushed to the printer driver cache and then the plaintext chunk in the server memory is destroyed immediately.

[0055] Physical Output: The printer receives a data stream containing traceability information and performs physical printing. Because the data stream is encrypted for a specific Device ID, even if an attacker intercepts the network packets, they cannot reconstruct the image on other printers.

[0056] Through the above embodiments, this system successfully achieved stable transmission of 10GB-level inkjet files, with preview response time controlled within 200ms. Furthermore, security audits verified that no complete plaintext source file remnants were found even when the client experienced network outages, power outages, or malicious attacks. Experimental results show that this system, while ensuring that GB-level image data remains on the ground throughout the entire process, improves transmission efficiency by approximately 40% (compared to traditional FTP) and reduces preview latency to millisecond levels, effectively verifying the dual technical advantages of this invention in terms of security and high performance.

Claims

1. A method for secure closed-loop transfer and physical output control of ultra-large raster image data, characterized in that, Includes the following steps: (1) Adaptive access and fragmentation steps: The client obtains the ultra-large raster image file to be processed, monitors the bandwidth jitter rate and packet loss rate of the current network link in real time, and calculates the optimal dynamic fragmentation granularity accordingly; the image file is logically cut into multiple data fragments, the hash feature value of each fragment is calculated, and a metadata index containing fragmentation order and verification information is established. Then, the fragmented data is asynchronously transmitted to the server through an encrypted channel. (2) Fragmentation and reassembly and encrypted storage steps: The server receives the fragmented data and verifies its integrity. After the verification is successful, it does not restore it to the original file, but uses a distributed storage strategy to persistently store the fragmented data. At the same time, it generates a global key that is uniquely bound to the image file, performs secondary encryption on the stored fragmented data, and generates a corresponding ciphertext index table. (3) Zero-buffered preview steps based on viewport mapping: In response to the preview request initiated by the client, the server constructs multi-resolution hierarchical data based on the image pyramid model and calculates the set of image slices corresponding to the current visible area according to the client's viewport coordinates; the server only extracts the encrypted slices corresponding to the set, dynamically decrypts and synthesizes digital watermarks in real time before transmitting them to the client. The client receives and renders the slice data in volatile memory, and releases the memory immediately after the preview ends, ensuring that the source file is not written to the ground throughout the entire preview process; (4) Streaming decryption and hardware binding output steps: In response to the physical printing command, the server locks the hardware fingerprint information of the target physical printing device and builds an end-to-end secure output channel. The server reads the stored encrypted fragments, performs streaming decryption and splicing of the image data according to the buffer size required by the print driver, dynamically embeds the production traceability QR code during the rasterization process, injects the processed data stream into the printer buffer in real time for physical output, and destroys the decrypted data immediately after output is completed.

2. The method as described in claim 1, characterized in that, In the adaptive access and fragmentation step, the method for calculating the optimal dynamic fragmentation granularity is as follows: The base fragmentation size is set to S. base Get the current real-time network bandwidth B and packet loss rate P; Define a fragmentation adjustment factor α, where α is directly proportional to the bandwidth B and inversely proportional to the packet loss rate P; Calculate the dynamic fragment size S dynamic = S base × α; When the network status fluctuates drastically and exceeds a preset threshold, the client adjusts the S of subsequent untransmitted fragments in real time during transmission. dynamic The value is updated, and the metadata index is updated.

3. The method as described in claim 1, characterized in that, In the zero-buffered preview step based on viewport mapping, the image slice set is obtained as follows: the server pre-downsamples the ultra-large raster image to generate N layers of different resolutions, each layer being divided into slices of fixed size by meshing; the client obtains the current zoom level and viewport center coordinates based on user interaction; the client matches the corresponding layer L according to the zoom level. i Based on the viewport center coordinates and the display area size, the required range of slice row and column numbers (R) to cover the area is calculated. min , R max ) and (C min C max The client only requests slice data within this range from the server.

4. The method as described in claim 1, characterized in that, In the streaming decryption and hardware binding output steps, the construction of the secure output channel includes: the server obtaining the unique hardware identifier (Device ID) of the target printer; the server using the hardware identifier, the timestamp of the current task, and the global key to generate a one-time printing session key using a hash algorithm; the server using the session key to establish an encrypted pipeline with the printing agent service, ensuring that the data stream transmitted to the printer can only be recognized and processed by this specific hardware device, preventing man-in-the-middle attacks or data replay to other devices for printing.

5. The method as described in claim 1, characterized in that, It also includes the steps for determining whether to resume interrupted transmission and transmit instantly: before the client uploads the fragment, it sends the hash feature value of the fragment to the server; the server searches the encrypted index table, and if it finds that the same hash feature value already exists in the system, it directly maps the stored data block to the metadata of the current image file and notifies the client to skip the transmission of the fragment. If an interruption occurs during transmission, the client records the fragment indexes that have been successfully transmitted. After the network is restored, the client will re-initiate a transmission request only for the dynamic fragment size corresponding to the fragment indexes that have not been transmitted.

6. A secure closed-loop transfer and physical output control system for ultra-large raster image data, characterized in that, include: Client interaction module: Provides a user interface for selecting files, setting printing parameters, and performing dynamic file segmentation calculations and uploading tasks for large files; It includes a memory-based rendering engine for receiving image slice streams sent from the server and displaying them without caching; The core server-side processing module receives fragmented data uploaded by the client, performs hash verification, distributed storage, and secondary encryption operations; it also includes an image processing engine for generating image pyramid levels and dynamically adding digital watermarks. Security control module: Used to manage global keys and permission policies, responsible for generating one-time session keys bound to the printer hardware fingerprint during the physical printing stage, and monitoring the decryption and destruction process of the data stream; Physical output driver module: Deployed on the print server, it receives streaming decrypted data from the server, converts the data into printer instruction set (PJL / PCL), and drives the physical printing engine to complete the final image output.

Images