Depth Image Feature Point Extraction Using Center of Gravity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for searching feature points in images with depth information, such as human recognition technologies, require extensive processing time and involve moving and resizing sliding windows, which is inefficient.
Innovation Solution
An image processing apparatus and method that uses a depth image to extract feature points by initializing a center of gravity and sequentially processing temporary and base lists to determine feature points based on surface scores and depth differences, reducing the time required to estimate the overall structure of an object like a human body.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional sliding window method is used to search for feature points, then comprehensive search coverage is achieved, but execution time becomes excessively long
Solution Approach 1:
The patent performs preliminary actions by first extracting the center of gravity of the object and using it as a starting point for feature point search. This preliminary positioning allows the algorithm to focus search efforts in relevant regions rather than exhaustively scanning the entire image, thereby reducing execution time while maintaining search completeness.
Solution Approach 2:
The patent segments the feature point search process into distinct stages: (1) center of gravity extraction, (2) temporary list generation based on depth information, (3) base list construction, and (4) feature point determination. This segmentation allows each stage to optimize for its specific task, improving overall efficiency compared to a monolithic sliding window approach.
2Measurement precision
If sliding window is moved to overall range with size changes, then thorough feature point detection is achieved, but processing complexity increases
Solution Approach 1:
The patent applies local quality by using depth information to identify regions with significant depth variations, which are more likely to contain feature points. Instead of uniformly processing the entire image, the algorithm focuses computational resources on local regions with high depth gradients, reducing processing complexity while maintaining detection thoroughness.
Solution Approach 2:
The patent performs preliminary analysis of depth information to identify candidate regions before conducting detailed feature point detection. By pre-processing the depth data to highlight areas of interest, the algorithm reduces the complexity of subsequent detection steps while ensuring thorough feature point identification.
3Measurement precision
If depth information is used for feature point extraction, then recognition accuracy is improved, but processing time increases
Solution Approach 1:
The patent performs preliminary processing of depth information to extract the center of gravity and generate a temporary list of candidate points. This preliminary extraction reduces the amount of depth data that needs to be processed in subsequent steps, allowing the algorithm to maintain high recognition accuracy while reducing overall processing time.
Solution Approach 2:
The patent segments the depth information processing into distinct phases: (1) center of gravity calculation from depth data, (2) temporary list generation using depth gradients, and (3) final feature point determination. This segmentation allows depth information to be used efficiently at each stage, improving recognition accuracy without proportionally increasing processing time.
Data Source
AI summary
An image processing apparatus for searching for a feature point by use of a depth image and a method thereof are provided. The image processing apparatus includes an input unit configured to input a three-dimensional image having depth information, a feature point extraction unit configured to obtain a designated point from an object image extracted from the depth image to obtain a feature point that is located at a substantially farthest distance from the designated point, and to obtain other feature points that are located at substantially farthest distances from feature points that are previously obtained as well as the designated point. The apparatus includes a control unit configured to control the input unit and the feature point extraction unit so that time in estimating a structure of the object is reduced, and a recognition result is enhanced.