Branch Structure Discriminator for Rotated Pattern Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing pattern recognition methods, such as those using cascade-connected weak discriminators, struggle to efficiently identify patterns that undergo significant in-plane and depth rotations, leading to increased processing time and reduced accuracy.
Innovation Solution
An information processing apparatus with a branch structure that learns and determines the optimal branch structure for a discriminator, using a preliminary learning unit to establish a preliminary discriminator for each variation category and a main learning unit to refine the main discriminator based on discrimination processing results, allowing for efficient detection across various rotations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If a cascade-connected weak discriminator structure is used for pattern recognition, then processing speed is improved through early termination, but accuracy deteriorates when detecting patterns with significant rotations
Solution Approach 1:
The detection process is segmented into multiple stages, where each stage contains cascade-connected weak discriminators. The segmentation allows early termination for non-target patterns while maintaining accurate detection capability through progressive refinement across stages, resolving the contradiction between speed and accuracy.
Solution Approach 2:
The system dynamically adjusts the number of weak discriminators and cascade depth at each stage based on the specific detection task. This dynamic configuration allows the system to optimize processing speed for simple cases while maintaining accuracy for complex rotated patterns, preventing the accuracy deterioration that would occur with a fixed rigid structure.
2Manufacturing precision
If the number of weak discriminators is increased to improve detection accuracy, then manufacturing precision of the discriminator improves, but device complexity increases
Solution Approach 1:
Instead of using a single complex discriminator with many weak discriminators, the system segments the detection into multiple stages, each with a manageable number of weak discriminators. This segmentation achieves high overall accuracy while keeping each individual stage's complexity controllable and manageable.
Solution Approach 2:
The system transitions from a one-dimensional linear cascade structure to a multi-dimensional staged structure. By adding the stage dimension, the system can distribute weak discriminators across multiple levels, achieving high accuracy without concentrating all complexity in a single linear path.
Data Source
AI summary
An information processing apparatus and method enables a pattern discriminator to learn. The apparatus establishes a branch structure appropriate for learning a discriminator having the branch structure without increasing processing time. The apparatus includes a preliminary learning unit to learn a preliminary discriminator for a respective one of a plurality of combinations of variations in variation categories in a discrimination target pattern. A branch structure determination unit is provided to perform discrimination processing using the preliminary discriminator and to determine a branch structure of a main discriminator based on a result of the discrimination processing. A main learning unit is included to learn the main discriminator based on the branch structure.