Hand Region Extraction in Distance Images for HMDs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods struggle to accurately extract hand regions from distance images, especially when other objects are present near the hands, making it difficult to operate head-mounted display (HMD) apparatuses and perform tasks like maintenance checkups in plant settings.
Innovation Solution
A method involving a region extraction device that divides a distance image into regions by generating pixel groups based on pixel value differences and positional relationships, using a processor to determine whether to merge regions and select those that satisfy predetermined conditions, thereby extracting hand regions even in the presence of nearby objects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing methods are used to extract hand regions from distance images, then the extraction process is simple, but the accuracy deteriorates when other objects are present near the hands
Solution Approach 1:
The patent segments the distance image into multiple regions by comparing pixel values between adjacent pixel pairs. This segmentation approach divides the image processing task into manageable parts, allowing accurate identification of hand regions even in the presence of other objects by analyzing local pixel value differences and merging adjacent regions systematically.
Solution Approach 2:
The patent changes the parameter used for region extraction from simple threshold-based methods to pixel value difference analysis. By calculating the absolute difference between pixel values of adjacent pixels and comparing it to a threshold, the system adapts to varying lighting and environmental conditions, improving extraction accuracy without requiring complex preprocessing.
2Reliability
If simple threshold-based methods are used for region extraction, then the processing speed is fast, but the reliability deteriorates in complex environments with multiple objects
Solution Approach 1:
The patent processes the distance image by segmenting it into regions through systematic comparison of adjacent pixel pairs. This segmentation approach maintains processing efficiency by using simple threshold comparisons while improving reliability through the systematic merging of adjacent regions based on pixel value differences, making the method robust to complex environmental conditions.
Solution Approach 2:
The patent introduces an intermediary step of calculating pixel value differences as a mediator between the raw distance image and the final extracted regions. This intermediary calculation layer transforms the raw pixel data into a form that is more suitable for reliable region extraction, enabling the system to distinguish hand regions from other objects even in complex environments.
3Measurement precision
If detailed region division is performed to improve extraction accuracy, then the measurement precision improves, but the processing time increases
Solution Approach 1:
The patent performs detailed region division by systematically processing adjacent pixel pairs and merging regions based on pixel value differences. This segmentation approach achieves accurate hand region boundaries by examining local pixel variations without requiring exhaustive analysis of the entire image, thus maintaining processing efficiency while improving boundary accuracy.
Solution Approach 2:
The patent applies partial action by focusing the detailed region division only on areas where pixel value differences indicate potential region boundaries. Rather than performing exhaustive analysis on the entire image, the method selectively processes adjacent pixel pairs that show significant value changes, achieving accurate boundaries with reduced processing time.
Data Source
AI summary
A method of extracting a region in a distance image including pixels, the method includes: for each of adjacent pixel pairs in the distance image, generating a third pixel group that includes a first pixel group to which a first pixel belongs and a second pixel group to which a second pixel belongs based on a difference between pixel values of the first pixel and the second pixel included in the adjacent pixel pair; dividing the distance image into regions by determining whether to generate a third region represented by the third pixel group by merging a first region represented by the first pixel group and a second region represented by the second pixel group, based on a positional relationship of points represented by pixels included in the third pixel group; and selecting a region that satisfies a predetermined condition from among the regions.