基于经验与灰度共生矩阵的花蛤辨识方法

为提高花蛤挑拣的效率，加快水产产业自动化生产步伐，提出一种基于经验和灰度共生矩阵的花蛤筛选方法。本方法属于无接触式辨识，在对工业相机采集到的图像进行预处理的基础上，截取并旋转单个目标图像，规范图像性质，采用统计学和灰度共生矩阵方法，提取花蛤或石头的纹理和形状等相关特征，运用支持向量机进行分类训练和测试。实验结果表明，本文所提出的方法可实现位置精准确定，选取的特征能较好地表示花蛤或石头，分类器算法高效简单，最终花蛤筛选识别率可达99%，可满足工业需求。

Clam Identification Based on Experience and Gray Level Co Occurrence Matrix

To improve the efficiency of clam selecting to accelerate the automated production in the aquatic industry,a clam screening is proposed based on experience and gray level co occurrence matrix as a non contact identification approach.After preprocessing the image collected by the industrial camera,target images were intercepted and rotated one by one,image properties normalized,features of clams or stones such as textures and shapes extracted using statistics and gray level co occurrence matrix,and finally training and testing done for classification using support vector machine.The experimental results show that the classification method proposed can accurately locate the clam position,well represent the clams or stones by the features selected,have efficient and simple classifier algorithm on work,and ensure the final clam screening and recognition rate of 99%,which meets the industrial needs.