线性变换结合拉普拉斯算子检测菌落方法的研究  

Research on Linear Transformation Combined with Laplace Operator for Colony Detection

作　　者：范玮 张新 李琰 成花 FAN Wei;ZHANG Xin;LI Yan;CHENG Hua(Hefei University of Economics,Hefei 230031,China)

机构地区：[1]合肥经济学院,安徽合肥230031

出　　处：《佳木斯大学学报（自然科学版）》2024年第10期35-38,共4页Journal of Jiamusi University：Natural Science Edition

基　　金：2022安徽省高校科研重点项目(2022AH052624)。

摘　　要：利用灰度图线性变换结合拉普拉斯算子检测发酵面制食品污染菌落。采集面包、馒头的霉菌图像经灰度化、高斯滤波等预处理,以函数imadjust将图像拉伸处理后再采用二阶导数拉普拉斯算子进行边缘提取,结果处理图像边缘清晰。比较一阶算子、Canny算子以及Log算子,线性拉伸处理后各个算子边缘提取结果不尽相同差异较大;以形态学腐蚀处理方法可以获得准确的菌落计数。Using grey graph linear transformation combined with Laplace operator to detect bacterial colonies in fermented food contamination.The mold images of bread and Mantou were collected and preprocessed by graying,Gaussian filtering,etc.The image was stretched by function imadjust,and then the second derivative Laplace operator was used for edge extraction.The results showed that the edges of the processed images were clear.Comparing first derivative operator,Canny operator,and Log operator,it is found that the edge extraction results of each operator after linear stretching processing are different and have significant differences.Morphological erode treatment methods can obtain more accurate colony counts.

关 键 词：线性拉伸 拉普拉斯算子 边缘提取 形态学腐蚀 菌落 

分 类 号：TP29[自动化与计算机技术—检测技术与自动化装置]