机电工程技术2024，Vol.53Issue(2)：35-38,88,5.DOI:10.3969/j.issn.1009-9492.2024.02.007
基于频谱遮罩的玻璃量器图像处理与围线检测方法
Image Processing and Scribe Detection Method for Glass Gauge Based on Spectrum Masking
摘要
Abstract
In order to solve the problem of excessive noise in edge detection of glass gauge scribe,an image preprocessing technology with a frequency domain mask created based on the spectral distribution characteristics of the scribe,and a scribe detection method based on image horizontal projection grayscale,are studied.Specifically,the feature spectrum of target is analyzed.A mask model and transfer function are established.The influence of function parameters on the mask processing results is discussed.And based on the different orientations of grayscale feature of the glass gauge scribe and the bottle body in the spatial domain,a method to detect the peak valley difference between the front and rear half scribe is proposed using the change rate of horizontal projection grayscale.By comparing the results of edge detection method and projection grayscale method in detecting scribe,the following conclusion can be drawn:the frequency domain mask processing technology can effectively blur and weaken the glass gauge and other interference signals,while retaining the scribe edges well without losing the size of valley peak difference,then greatly reducing the difficulty of subsequent image processing.The projection grayscale method can effectively detect the highest point of the up edge and the lowest point of the down edge,and the method has strong robustness to frequency domain masking processing.关键词
玻璃量器/图像处理/频域/遮罩图像/围线检测Key words
glass gauge/image processing/frequency domain/mask image/scribe detection分类
信息技术与安全科学引用本文复制引用
蔡永洪,马婷婷,杨叶花..基于频谱遮罩的玻璃量器图像处理与围线检测方法[J].机电工程技术,2024,53(2):35-38,88,5.基金项目
广东省市场监督管理局科技项目(2020CJ02) （2020CJ02）
