Fast Image Data Storage Using Bitwise Masking for Security Documents
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for storing high-resolution image data in security document identification systems are too time-consuming, exceeding the required 15ms storage time, which can lead to system breakdowns and reduced identification speed or functionality.
Innovation Solution
A fast storage method that converts single-byte image data into long integer type data using bitwise operations with data masks to extract and encode image data, reducing the number of operations and enabling faster compression and storage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional point-by-point storage method is used, then complete image information can be stored, but storage time exceeds the required 15ms limit
Solution Approach 1:
The patent segments the image data into multiple 8x8 pixel blocks, processing and storing each block independently through bitwise operations. This segmentation enables parallel processing of multiple blocks simultaneously, dramatically reducing total storage time while preserving all image information through systematic encoding of each segment.
Solution Approach 2:
The patent changes the data representation parameter from individual pixel storage to block-based bitwise encoding. By representing 8x8 pixel blocks using bitwise AND, OR, and NOT operations with predefined masks, the storage operation transforms from sequential pixel-by-pixel writing to accelerated vectorized block processing, achieving sub-15ms storage time.
2Loss of time
If memcpy function is used for fast storage, then storage time is reduced to 18ms, but CPU cannot be interrupted causing system breakdown
Solution Approach 1:
The patent divides the large-scale image copy operation into multiple smaller 8x8 pixel block operations. Each block can be processed independently with interrupt capability, allowing the CPU to handle other tasks during storage operations. This segmentation maintains fast storage performance while restoring system responsiveness and stability.
Solution Approach 2:
The patent introduces bitwise mask operations as intermediaries between the source image data and storage destination. These mask-based encoding operations serve as a processing layer that enables efficient data transformation and compression while maintaining interruptibility, unlike direct memcpy which monopolizes CPU resources.
3Measurement precision
If high-resolution images are stored, then complete image detail is preserved, but storage time increases beyond 15ms
Solution Approach 1:
The patent segments high-resolution images into multiple 8x8 pixel blocks, processing each block through efficient bitwise operations. This segmentation allows parallel processing of numerous small blocks simultaneously, maintaining complete image detail while reducing total storage time to under 15ms through optimized block-level operations.
Solution Approach 2:
The patent changes the storage parameter from full-resolution pixel-by-pixel writing to block-based bitwise encoding. By using predefined masks and bitwise operations on 8x8 blocks, the system preserves complete image information while achieving accelerated storage performance that meets the 15ms requirement.
Data Source
Figure 1
Figure 2~3
Figure 4a~4c
AI summary
Disclosed are a valuable-file identifying method and an identifying device thereof. The identifying method and the identifying device store image data of a valuable file using a fast storage method for image data. The fast storage method for image data comprises: compulsively converting collected single-byte image data into long-integer image data; using N data masks which correspond to each other through a "bitwise AND" operation to extract the long-integer image data in such a manner that N points are extracted from M * N points in each line and one point is extracted from L points in each column, where N is an integer greater than or equal to 2, L and M are all integers greater than or equal to 1; and integrating the data extracted respectively by N data masks through a "bitwise OR" operation to obtain coded image data and store same. In conclusion, the present invention converts single-byte image data into long-integer image data, and extracts multipoint data at one time using a plurality of data masks which correspond to each other, reducing the number of operations, and achieving fast, compressed storage of images.