卷积神经网络在图像分类上的应用综述.pdf

Artificial Intelligence and Robotics Research 人工智能与机器人研究, 2018, 7(1), 17-24 Published Online February 2018 in Hans. /journal/airr /10.12677/airr.2018.71002 Summary of Application of Convolution Neural Network on Image Classification Zeming Yang, Jun Liu, Cheng Xue, Zihong Yu Beijing Wuzi University, Beijing nd st th Received: Feb. 2 , 2018; accepted: Feb. 21 , 2018; published: Feb. 28 , 2018 Abstract In recent years, image classification occupies an important position in pattern recognition. The rapid growth of image data for image information analysis and processing put forward new re- quirements. Convolution neural network came into being, with its powerful image recognition classification ability is widely used in a variety of image classification system, and achieved very significant results. Firstly, this paper reviews the development of image classification technology, and introduces the classification methods of feature space and feature space and the image fea- ture extraction methods. Secondly, the research of convolution neural network on image classifi- cation is introduced. Finally, the problems existing in the image classification of convolution neur- al networks are summarized, and the future development direction is summarized. Keywords Image Classification, Convolution Neural Network, Image Feature Extraction Method 卷积神经网络在图像分类上的应用综述 杨泽明，刘 军，薛 程，于子红 北京物资学院，北京 收稿日期：2018年2月2 日；录用日期：2018年2月21 日；发布日期：2018年2月28 日 摘 要 近些年来，图像分类在模式识别领域占据着重要的地位。迅速增长的图像数据对于图像信息的分析与处 理提出了新的要求。卷积神经网络应运而生，以其强大的图像识别分类能力被广泛的应用于各种图像分 文章引用: 杨泽明, 刘军, 薛程, 于子红. 卷积神经网络在图像分类上的应用综述[J]. 人工智能与机器人研