A Classification Method of Circular Objects Based on Gray Level Statistical Features

This paper presents a classification method of circular objects based on gray statistical features. Firstly,use circular object detection to extract the region where the circular object is located. Then, standardize the detected circular region to the circular region with the uniform radius. At the same time, normalize the gray level of the circular region; then, we divided the standard circular region into little regions according to the determined pattern, and extract the gray level statistical features of each sub-region , which are combined into feature vectors. In this paper, we use Support Vector Machine and K-Nearest Neighbor for classification experiments.Experiments show that the method has a good classification effect.

Circular object image, image classification, feature extraction , image region division

Hanxu Liu, Mianshu Chen, A Classification Method of Circular Objects Based on Gray Level Statistical Features. 2019 Scientific Conference on Network, Power Systems and Computing (NPSC 2019), 2019: 136-139. DOI: https://doi.org/10.33969/EECS.V3.031.