基于图像纹理特征的土鸡蛋微裂纹无损检测

为了提高土鸡蛋表面微裂纹检测的准确度和效率,提出一种基于图像纹理特征的土鸡蛋微小裂纹无损检测方法.利用工业相机对150枚土鸡蛋采集数字图像,采用高斯滤波、灰度变换等方法对土鸡蛋图像进行预处理;利用灰度共生矩阵进一步提取图像纹理特征,将纹理特征参数作为不同分类器包括簇类独立软模式法、线性判别分析(linear discriminant analysis,LDA)和偏最小二乘支持向量机输入,进行土鸡蛋有无裂纹判别.结果表明,采用图像纹理特征参数建立的土鸡蛋有无裂纹LDA模型判别准确度最高,达到96.0%.

Nondestructive method to detect small crack of native egg based on image textural feature

In order to improve the detection accuracy and efficiency for shell small crack of native egg, a method based on textural feature of images was proposed. Digital image was obtained by industrial camera and then preprocessed by Gauss filtering and gray-scale transformation. After that, textural feature of images was extracted based on gray level co-occurrence matrix (GLCM). For crack identification, textural feature parameters were used as classifier input which included soft independent modeling of class anal-ogy ( SIMCA) , linear discriminant analysis ( LDA) and least squares support vector machine ( LS-SVM) . The results showed that LDA model achieved the best correct classification rate ( CCR) at 96%.