CAJ | 학술논문

采用机器视觉获取了西瓜子的面积、周长、最小外接矩和圆形度等外形特征，而后使用遗传算法优化的最小二乘支持向量机算法对西瓜子外观品质进行分类识别，最终实现了破损瓜子、普通瓜子和优质瓜子的区分。实验结果表明，基于最小二乘支持向量机分类的西瓜子外形检测方法能够很好地实现西瓜子外观品质的识别检测。
채용궤기시각획취료서과자적면적、주장、최소외접구화원형도등외형특정，이후사용유전산법우화적최소이승지지향량궤산법대서과자외관품질진행분류식별，최종실현료파손과자、보통과자화우질과자적구분。실험결과표명，기우최소이승지지향량궤분류적서과자외형검측방법능구흔호지실현서과자외관품질적식별검측。
A machine vision system is developed to investigate the detection of watermelon seeds exterior quality. The main characteristics of watermelon seeds appearance including area, perimeter, roughness and minimum enclosing rectangle are calculated by image analysis. Least Square Support Vector Machine optimized by genetic algorithm is applied for the classification of watermelon seeds exterior quality, and the broken melon seeds, normal melon seeds and high-quality seeds are distinguished finally. The experimental results show that the watermelon seeds exterior quality can be well detected and classified by machine vision based on least squares support vector machine.