Image Correction Apparatus Using Line Memory for Optical Distortion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image correction technologies require large amounts of main memory and cache memory to correct optical distortion, which is costly and inefficient, especially in image pickup apparatuses where memory access involves significant computation and resource constraints.
Innovation Solution
An image correction apparatus that utilizes a line memory for storing image data and a distortion correction value storage to select and correct pixels in fixed-size regions, minimizing memory usage through pixel interpolation and efficient memory access, thereby reducing the required size of both main and cache memory.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional distortion correction methods are used, then optical distortion is corrected, but large amounts of main memory and cache memory are required
Solution Approach 1:
The patent divides the image data into fixed-size blocks (e.g., 8x8 pixels) and processes each block independently for distortion correction. This segmentation allows the system to use a line memory with a predetermined number of lines (e.g., 16 lines) rather than requiring large cache memory, as only the necessary lines for each block are stored and processed sequentially.
Solution Approach 2:
The patent pre-calculates and stores distortion correction values in a distortion correction value storage unit before actual image processing. These correction values are computed in advance based on the distortion characteristics of the optical lens, allowing the correction process to proceed efficiently without requiring large memory resources during real-time processing.
2Reliability
If large cache memory is allocated for distortion correction, then correction quality is maintained, but computation cost and manufacturing expense increase
Solution Approach 1:
By segmenting image processing into fixed-size blocks and using a line memory structure, the patent eliminates the need for large cache memory allocations. This reduces manufacturing costs while maintaining correction quality, as the system only stores the minimum necessary data (predetermined number of lines) in memory at any given time.
Solution Approach 2:
The patent uses a simplified memory structure (line memory with predetermined lines) that copies only the essential data needed for distortion correction, rather than maintaining large cache memory copies of entire image frames. This approach reduces memory capacity requirements and manufacturing expenses while preserving correction accuracy.
3Productivity
If memory access is increased for distortion correction, then correction processing is performed, but computation cost and processing time increase
Solution Approach 1:
The patent processes image data in fixed-size blocks sequentially, accessing only the predetermined number of lines in the line memory for each block. This segmented approach minimizes memory access operations compared to processing entire frames, reducing computation time and memory access overhead while maintaining correction capability.
Solution Approach 2:
Distortion correction values are pre-calculated and stored before image processing begins. During actual correction, the system only needs to retrieve these pre-computed values and apply them to each block, significantly reducing computation time and memory access requirements compared to calculating corrections in real-time.
Data Source
Figure 1
Figure 2A~2C
Figure 3A~3C
AI summary
A line memory (10) stores image data captured through optical lenses. A distortion correction value storage (3) stores distortion correction values used in the correction of optical distortion occurring in the image. An image data control unit (8) selects pixels of image data stored in the line memory (10) as pixels to be corrected. Then the image data control unit (8) determines a region for use in correction, having the fixed number of pixels horizontally and the fixed number of lines vertically. Then the image data control unit (8) reads out the image data from the line memory (10). A distortion correction unit (11) determines pixel values for the pixels to be corrected, based on (i) the pixels for use in correction, which are contained in the image data, within the region for use in correction, and (ii) the distortion correction values.