基于计算机视觉的苹果在线高效检测与分级系统

摘要

The system of online automatic detection and classification of apple was carried out corperating with the enterprise ,based on CCD cameras , sensors , computer and MATLAB.A construction parameter model of three imaging system and a probability model of the apples reached steady-state were established.The color characteristics ratio law and fuzzy clustering were used to detect the defect area quickly and accurately.A new algorithm for defect area by which the curvature of the surface area of the reasonable impact of the conversion was resolved was proposed.Experimental results prove the feasibility and accuracy of the system.%基于由计算机、摄像机、传感器为主体构建的硬平台和MATLAB为主的软平台,与企业合作研发出苹果在线自动检测、分级系统.推导出摄像机合理位姿参数计算式,建立了苹果从动态向稳态过渡的数学模型.提出一种检测果形和尺寸方法,使在线动态检测能够实现.依据苹果色泽特征比率规律,结合模糊聚类知识实现对缺陷域快速、准确的识别.提出一种缺陷面积的新算法,解决了表面曲率对合理换算面积的影响问题.试验表明所研发的系统能实现高效、准确的检测分级.