基于机器视觉的脱绒棉种染色特征的提取算法  

The Extract Algorithms of Staining Characteristics of Cottonseeds Based on Machine Vision

作　　者：董翠翠[1] 李景彬[1] 坎杂[1] 王晓华[1] 王晨[1] 彭顺正 

机构地区：[1]石河子大学机械电气工程学院,新疆石河子832000

出　　处：《农机化研究》2015年第11期51-54,共4页Journal of Agricultural Mechanization Research

基　　金：国家自然科学基金项目(31260290)

摘　　要：针对种子活力四唑染色法检测中由于人工视觉疲劳、劳动强度大等因素引起的人工误判情况,提出了一种基于机器视觉的脱绒棉种染色特征的提取算法。此算法是在随机选取了庄稼汉品种的脱绒棉种进行四唑试验后进行的,通过对染色棉种进行图像采集,对图像进行灰度变换、HSI颜色空间转换、中值滤波、自动阈值分割等预处理,经去噪、补洞后得到整粒脱绒棉种的面积和染色部分的面积。同时,计算出图像染色比例,为脱绒棉种活力在线检测奠定理论基础。The situation of manual misjudged was caused by artificial visual fatigue , high labor intensity .A extraction al-gorithm of Cottonseeds ’ staining characteristics is proposed based on machine vision for avoid this situation .This algo-rithm is applied to cottonseeds which named Zhuangjiahan , and these cottonseeds were tetrazolium test .The image acqui-sition were from staining cottonseeds , then these images were preprocessed by gray-scale transformation , HSI color space conversion , median filtering , automatic threshold segmentation etc .The whole cottonseeds area and staining section area got by de-noising , fill holes .The ratio of the staining image calculated for the theoretical basis of viability online testing of cottonseeds .

关 键 词：脱绒棉种 活力 四唑试验 机器视觉 图像处理 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] S722.14[自动化与计算机技术—计算机科学与技术]