西红柿果实目标识别方法研究——基于模糊聚类算法  

Study on Identification of Tomatos——Based on Fuzzy Clustering Algorithm

作　　者：王富春[1] 徐越[2] 宋怀波[2] 

机构地区：[1]柳州职业技术学院机电工程系,广西柳州545006 [2]西北农林科技大学机械与电子工程学院,陕西杨凌712100

出　　处：《农机化研究》2015年第10期24-28,33,共6页Journal of Agricultural Mechanization Research

基　　金：国家"863计划"项目(2013AA10230402);陕西省自然科学基金项目(2014JQ3094);西北农林科技大学本科生优质课程建设项目(2012)

摘　　要：为了准确识别成熟的西红柿目标,提出了一种模糊C-均值聚类算法(Fuzzy Clustering Means,FCM)的西红柿目标分割方法。该方法首先利用FCM算法对西红柿图像进行模糊聚类,并对聚类后的果实图像与丢失的部分目标图像进行相加,以得到更加完整的西红柿目标;然后对西红柿目标进行二值化、去噪、开运算与闭运算等处理,完成西红柿目标的分割。为了验证算法的有效性,利用20幅图像进行了试验并与K-means算法和Otsu算法分割效果进行了对比。结果表明:利用文中算法所分割出的西红柿目标最高分割误差率均低于Kmeans算法和Otsu算法,平均分割错误率为1 6.5 5%,比K-means算法低了3.5 6%,比Otsu算法低了1 2.8 0%。这表明,将该方法应用于西红柿目标的识别是可行的。In order to identify tomatoes accurately , an identification method of tomatoes from natural scenes based on fuzzy clustering algorithm was presented .First, tomato image was clustered by using FCM algorithm .Then, miss-clus-tered target was added to fruit image after clustering , which could help to get a more complete target .Finally , binaryza-tion , de-noising , and morphology operations such as image open and image close were used to complete the segmentation of tomato targets .In order to verify the validity of this algorithm , a test was conducted by using proposed algorithm with 20 tomatoes images and compared with K-means algorithm and Otsu algorithm respectively .The experiment results show that the highest error segmentation rate of the proposed algorithm was lower than those of K -means algorithm and Otsu al-gorithm, the average error rate of the proposed algorithm was 16.55%,3.56% less than that of the K-means algorithm and 12 .80%.less than that of the Otsu algorithm.In conclusion , the proposed algorithm is feasible in recognizing toma -toes.

关 键 词：西红柿 FCM算法 图像分割 目标识别 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术] S641.2[自动化与计算机技术—计算机科学与技术]