Resistor Detection Based on Convolutional Neural Networks
Resistor detection are mostly suffered from compact size and multiple interferences of environment.In this paper,a method for resistor detection was proposed,which combined selective search(SS),convolutional neural networks(CNNs)and support vector machine(SVM).Using improved selective search method to reduce the time of generating candidate regions; taking advantages of independent features extraction and autonomous learning of CNNs,and avoiding the process of complicated features extraction manually; fewer training samples and parameters are needed with this method.According to the parameter tuning and experiment validation,this method achieved a precision of 96.15%on testing dataset established by ourselves.In addition,this method is also applicable to the identification of other small components.
selective search convolutional neural networks support vector machine resistor detection
Chun Liu Yudeng Shi
School of Computer,HuBei University of Technology Wuhan,China
国际会议
重庆
英文
91-94
2017-10-03（万方平台首次上网日期，不代表论文的发表时间）