基于灰度识别的猕猴桃形状疤痕图像纹理特征提取方法  

Texture feature extraction method of kiwifruit shape scar image based on gray level recognition

作　　者：纪娜[1] 何国荣[1] JI Na;HE Guorong(Yang Ling yocational and Technical College,Yangling Shanxi712100,China)

机构地区：[1]杨凌职业技术学院,陕西杨凌712100

出　　处：《自动化与仪器仪表》2019年第1期159-162,共4页Automation & Instrumentation

基　　金：2016年陕西省教育厅科研项目"基于机器视觉和重量检测的猕猴桃分选技术研究"(16JK1872);2016年杨凌职业技术学院校内自然科学研究基金"基于机器视觉的猕猴桃分级技术研究"(A2016031)阶段性成果

摘　　要：针对传统猕猴桃形状疤痕图像纹理特征提取方法运行效率低的问题,提出基于灰度识别的猕猴桃形状疤痕图像纹理特征提取方法。首先利用TUT-34rs传感器完成图像采集,同时采用邻域平均法清洗采集到的图像,然后在此基础上进行图像灰度化和二值化处理,并进行图像边缘检测,最后在上述一系列准备过程基础上进行猕猴桃形状疤痕图像纹理特征提取。结果表明:利用该方法对猕猴桃形状疤痕图像纹理特征进行提取,平均花费的时间要比传统方法少2. 88 s,由此证明该方法的运行效率要好于传统方法。Aiming at the inefficiency of traditional texture feature extraction methods for Kiwifruit shape scar images,a texture feature extraction method based on gray level recognition is proposed. Firstly,TUT-34 rs sensor is used to complete the image acquisition,and domain average method is used to clean the collected image. Then,image gray and binary processing are carried out on this basis,and image edge detection is carried out. Finally,texture feature extraction of kiwifruit shape scar image is carried out on the basis of the above-mentioned series of preparation process. The results show that the average time of extracting texture features from kiwifruit scar image is 2. 88 s less than that of traditional methods,which proves that this method is more efficient than traditional methods.

关 键 词：灰度识别 猕猴桃形状疤痕图像 纹理特征 特征提取 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]