Abstract:This paper introduces the characteristics of content-based image retrieval technology, and describes the edge features of medical images with histogram of oriented gradient features. Then these images are classified with support vector machine. The effect of several kernel functions on classification results is also discussed in this paper. This paper points out that the classification method based on HOG and SVM can help medical staff retrieve medical imags.
周显国. 基于梯度方向直方图特征和支持向量机的医学影像分类方法[J]. 中国医疗设备, 2014, 29(2): 132-133.
ZHOU Xian-guo. Classification Method for Medical Images Based on HOG Features and SVM. China Medical Devices, 2014, 29(2): 132-133.