Depth Image Edge Extraction via Multi-Algorithm Fusion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing depth image segmentation methods are complex and often fail to provide accurate and stable results due to noise, especially when using single frames or zero-value pixels.
Innovation Solution
The method involves calculating multiple sets of edge pixels using different edge detection algorithms and combining them to obtain a final set, which includes pixels exceeding a preset threshold or connected to the initial set, thereby enhancing accuracy and stability through fusion and compensation techniques.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple edge detection algorithms are used to calculate edge pixels, then segmentation accuracy is improved, but computational complexity increases
Solution Approach 1:
The patent combines multiple edge detection algorithms (Canny, Sobel, Laplacian) to detect edge pixels from depth images. By merging the results of different algorithms, the system achieves more accurate and robust edge detection compared to using a single algorithm, resolving the contradiction between accuracy and complexity through systematic integration.
Solution Approach 2:
The system performs iterative refinement where edge pixels detected by multiple algorithms are combined, and the results are used to refine the segmentation. The feedback mechanism allows the system to adjust and improve edge detection accuracy across multiple processing passes, balancing the increased computational complexity with progressive accuracy improvement.
2Stability of the object's composition
If multiple frames are fused to obtain depth image, then segmentation stability is improved, but processing time increases
Solution Approach 1:
The patent performs preliminary fusion of multiple depth frames before edge detection and segmentation processing. By pre-fusing the frames to create a stable depth image, the system reduces noise and improves segmentation stability in advance, preventing the need for repeated processing and ultimately reducing total processing time despite the initial fusion cost.
Solution Approach 2:
The system fuses a specific number of frames (excessive action) to ensure sufficient stability, rather than using just a single frame. This partial redundancy in frame fusion provides robustness against noise and temporal variations, achieving stable segmentation results that justify the additional processing time投入.
3Manufacturing precision
If edge pixels are compensated and refined through multiple processing steps, then segmentation precision is improved, but computational load increases
Solution Approach 1:
The patent segments the edge detection process into distinct stages: initial edge pixel detection by multiple algorithms, compensation of detected edges, refinement of edge pixels, and final segmentation. This segmentation of the processing pipeline allows each stage to focus on specific tasks, improving overall precision while managing computational energy through structured processing.
Solution Approach 2:
The edge compensation and refinement processes use the detected edge pixels themselves to improve their own quality. The system leverages the inherent information in the detected edges to compensate for missing or inaccurate pixels, reducing the need for external computational resources and achieving precision improvement through self-refinement.
Data Source
AI summary
The embodiments of the present disclosure disclose a method and apparatus for extracting an edge of an object in a depth image, a storage medium, and a computer device, to provide a solution which is simple to implement and accurate in edge extraction. The method for extracting an edge of an object in a depth image includes: calculating at least two sets of edge pixels of a first depth image using at least two edge detection algorithms; and acquiring a final set of edge pixels by combining the at least two sets of edge pixels. The solution according to the embodiments of the present disclosure is simple to implement, and may obtain a more accurate set of edge pixels, thereby obtaining an accurate pixel-level segmentation result.