Image Processing Device Block Matching Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing technologies face challenges in efficiently tracking moving objects across images due to high storage requirements and inefficient data transfer processes, particularly in systems with limited memory capacity.
Innovation Solution
The implementation of an image processing device that divides reference images into multiple regions, allowing for sequential data transfer and analysis, reducing storage needs and enhancing data transfer efficiency by processing image data in units of blocks within these regions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If reference images are stored in full for block matching processing, then tracking accuracy is improved, but storage capacity requirements increase
Solution Approach 1:
The reference image is divided into multiple divided regions, and only the currently needed region is loaded into the image processing device's storage at any given time. This segmentation allows the system to process large reference images without requiring the entire image to be stored simultaneously, thus reducing storage capacity requirements while maintaining tracking accuracy through sequential processing of all regions.
2Reliability
If entire reference images are transferred to processing device, then processing completeness is improved, but data transfer time increases
Solution Approach 1:
The reference image is segmented into multiple divided regions that are transferred to the image processing device sequentially rather than all at once. Each divided region is processed completely before the next region is transferred, ensuring processing completeness while significantly reducing the data transfer time at any given moment compared to transferring the entire reference image simultaneously.
Solution Approach 2:
The system employs periodic transfer and processing of divided regions in a cyclic manner. Each region is transferred, processed for block matching, and then the next region is transferred, creating a periodic workflow that balances data transfer time with processing completeness by maintaining a steady rhythm of data movement and computation.
3Productivity
If block matching is performed on entire reference image at once, then processing efficiency is improved, but memory bandwidth requirements increase
Solution Approach 1:
The block matching processing is performed on divided regions separately and sequentially rather than on the entire reference image simultaneously. This segmentation reduces the memory bandwidth requirements at any given time since only the data for one divided region needs to be accessed and processed at a time, while the overall processing efficiency is maintained through systematic handling of all regions.
Data Source
AI summary
An imaging device for tracking moving objects includes an image output device outputting first and second images that include a moving object to be tracked. A memory stores image data from the image output device. An image processing device includes local storage and performs a block matching processing comprising evaluating a first block in the first image, dividing the second image into multiple regions, storing image data for a divided region in the local storage unit, determining similarity of the first block to blocks in each of the divided regions in turn, and then determining the similarity of the first and second images according to results obtained for each of the divided regions.