Bug Image Keypoint Sequencing for Bent Posture Identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing trapped bug identification systems struggle with poor identification accuracy when bugs are in bent postures due to challenges in extracting morphological characteristics accurately.
Innovation Solution
A method and system that involves extracting bug data, imparting point and sequence information to at least three designated points on the bug, and using machine learning with trained models to enhance identification accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional bug identification systems are used, then the system is simple to operate, but the identification accuracy deteriorates when bugs are in bent postures
Solution Approach 1:
The bug image is segmented into multiple key points (at least three points) that represent critical morphological features. This segmentation allows the system to capture essential shape information without processing the entire complex image, thereby improving identification accuracy while managing data complexity.
Solution Approach 2:
The invention transitions from analyzing two-dimensional image pixels to extracting and analyzing point coordinate data with sequence information. By converting image data into a different dimensional representation (x, y coordinates with sequence ordering), the system achieves better posture invariance and identification accuracy.
2Measurement precision
If morphological characteristics are extracted from bent bugs, then identification can be performed, but the extraction accuracy deteriorates
Solution Approach 1:
The system performs preliminary extraction of key point coordinates and sequence information before the actual identification process. By pre-processing the bug image to extract essential geometric features (at least three points with sequencing), the system prepares robust data that maintains reliability even when bugs are in bent or overlapping postures.
Solution Approach 2:
The invention changes the parameters used for identification from pixel-based morphological characteristics to coordinate-based point information with sequence data. This parameter transformation makes the identification system more robust to posture variations, as point coordinates and their sequences remain reliable even when bugs are bent or overlapping.
3Measurement precision
If more bug features are analyzed, then identification accuracy improves, but the processing time increases
Solution Approach 1:
The system extracts only the essential features needed for identification - specifically at least three key points with their coordinates and sequence information - from the bug image. This selective extraction approach captures sufficient morphological information for accurate identification while avoiding the computational overhead of analyzing all image pixels or excessive features.
Data Source
AI summary
A method for generating bug identification image data according to the present invention is a method for generating bug identification image data in which pre-processing image data with an image of a bug is processed to generate the bug identification image data, and the method includes: a bug extracting step of extracting bug data on the bug from pre-processing image data; a point information imparting step of imparting point information data to at least three points designated from among the extracted bug data; and a sequence information imparting step of imparting sequence information on an order to the point information data at the at least three points designated in the point information imparting step.