Image Segmentation of Field Rape Based on Template Matching and K-means Clustering

[摘要] Giving that the changing light in the natural condition has negative impacts on the image segmentation of rape fields, the image of rape was processed by template match algorithm and K-means clustering algorithm to extract the rape flowers. In order to achieve the accurately segmentation of the rape flowers, firstly, creating a template library and using the template matching algorithm to locate the target area of the test image. Then, the processed image will be convert to LAB color space, and using K-means clustering algorithm to classify accurately again. Finally, the extracted rapeseed area will be processed by morphology operation. The experimental results indicate that this method can achieve the goal of extracting the rape flowers completely, and it is effectively to remove the negative impacts of the light.

[发布机构] School of Math and Computer, Wuhan Polytechnic University, Wuhan; 430023, China^1;Key Laboratory of Biology and Genetic Improvement of Oil Crops, Ministry of Agriculture, Oil Crops Research Institute, Chinese Academy of Agricultural Sciences, Wuhan; 430062, China^2

[效力级别] 无线电电子学 [学科分类] 计算机科学（综合）

[关键词] K;means clustering;K-Means clustering algorithm;Lab color space;Morphology operations;Natural conditions;Processed images;Template libraries;Template-matching algorithms