Classification Tree Pixel Selection for Multi-Source Image Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing pattern recognition methods using classification trees face low recognition performance and an exponential increase in tree structure size when combining data from multiple acquisition methods, making real-time processing and accurate classification impractical.
Innovation Solution
An information processing apparatus and method that acquires multiple image sets from different acquisition methods, generates partial image sets, and constructs a tree structure by selecting pixel positions and acquisition methods based on pixel values, allowing for efficient classification without increasing tree structure size.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If classification trees are created comprehensively applied to images obtained by all acquisition units, then recognition accuracy is improved, but device complexity increases enormously due to exponential tree size
Solution Approach 1:
The patent divides the classification process into multiple stages, creating separate classification trees for different acquisition units (luminance, range, color) rather than one comprehensive tree. Each tree handles a specific data type, and their results are combined later, preventing exponential complexity while maintaining accuracy
Solution Approach 2:
The patent merges the results from multiple separate classification trees through a combination process, integrating the classification outcomes from luminance, range, and color acquisition units to achieve accurate pattern recognition without creating a single exponentially large tree structure
2Measurement precision
If multiple image sets from different acquisition methods are processed, then recognition accuracy is improved, but execution speed decreases due to increased processing complexity
Solution Approach 1:
The patent segments the processing of multiple image sets by creating separate classification trees for each acquisition unit, allowing parallel or sequential processing of simpler trees rather than one complex tree, thereby maintaining execution speed while improving accuracy
Solution Approach 2:
The patent processes only the necessary portions of multiple image sets through specialized classification trees, avoiding full comprehensive processing of all data through a single tree, thus maintaining speed while achieving accurate classification through selective processing
3Device complexity
If a single classification tree is used, then device complexity is reduced, but recognition accuracy deteriorates compared to using multiple acquisition methods
Solution Approach 1:
The patent combines the results from multiple simple classification trees (each handling a specific acquisition unit) to achieve recognition accuracy comparable to or exceeding that of a single comprehensive tree, while keeping individual tree structures simple and manageable
Data Source
AI summary
An information processing apparatus including an acquisition unit that acquires a plurality of image sets obtained by different acquisition methods, a generation unit that generates partial image sets by extracting corresponding partial images from respective images of the image sets, an assignment unit that assigns the partial image sets to a root node of a tree structure, a setting unit that sets, at each node of the tree structure, positions of a plurality of pixels for each partial image set assigned to the node, a determination unit that determines whether any one of the plurality of pixels in a. partial image obtained by a predetermined one of the different acquisition methods in each partial image set has an invalid value, and a selection unit that selects art acquisition method based on the determination.