Research on Adaptive Edge Detection Method of Part Images Using Selective Processing

Visual quality inspection of part surfaces is a crucial step in industrial production. Image edge detection is a common technique for assessing the surface conditions of parts. However, current methods have limitations, including poor noise filtering, low adaptability, and inadequate accuracy of edg...

Full description

Saved in:
Bibliographic Details
Published in:Processes 2024-10, Vol.12 (10), p.2271
Main Authors: Li, Yaohe, Jin, Long, Liu, Min, Mo, Youtang, Zheng, Weiguang, Ge, Dongyuan, Bai, Yindi
Format: Article
Language:English
Subjects:
Accuracy
Adaptability
Adaptive algorithms
Algorithms
Data compression
Deep learning
Edge detection
Efficiency
Image acquisition
Image enhancement
Image filters
Image quality
Information processing
Inspection
Methods
Neural networks
Parameters
Plant layout
Production processes
Quality control equipment
Wavelet transforms
Citations: Items that this one cites
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Visual quality inspection of part surfaces is a crucial step in industrial production. Image edge detection is a common technique for assessing the surface conditions of parts. However, current methods have limitations, including poor noise filtering, low adaptability, and inadequate accuracy of edge detection. To overcome these challenges, this study proposes an adaptive edge detection method for part images using selective processing. Firstly, this method divides the input image into noise, edge, and noise-free blocks, followed by selective mixed filtering to remove noise while preserving original image details. Secondly, a four-parameter adaptive selective edge detection algorithm model is constructed, which adaptively adjusts parameter values based on image characteristics to address issues of missing edges and false detections, thereby enhancing the adaptability and accuracy of the method. Moreover, by comparing and adjusting the four parameter values, different edge information can be selectively detected, enabling rapid acquisition of desired edge detection results and improving detection efficiency and flexibility. Experimental results demonstrated that the proposed method outperformed existing classical techniques in both subjective and objective evaluations, maintaining stable detection under varying noise conditions. Thus, this method was validated for its effectiveness and stability, enhancing production efficiency in manufacturing processes of parts.
ISSN:2227-9717
2227-9717
DOI:10.3390/pr12102271