Down image recognition based on deep convolutional neural network

作     者：Wenzhu Yang Qing Liu Sile Wang Zhenchao Cui Xiangyang Chen Liping Chen Ningyu Zhang 

作者机构：School of Cyber Security and ComputerHebei UniversityBaoding 071002PR China 

出 版 物：《Information Processing in Agriculture》 (农业信息处理（英文）)

年 卷 期：2018年第5卷第2期

页      面：246-252页

核心收录：

学科分类：0710[理学-生物学] 0711[理学-系统科学] 0830[工学-环境科学与工程（可授工学、理学、农学学位）] 0907[农学-林学] 07[理学] 0908[农学-水产] 0905[农学-畜牧学] 0707[理学-海洋科学] 0906[农学-兽医学] 0829[工学-林业工程] 0901[农学-作物学] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

基　　金：supported by the Natural Science Foundation of Hebei Provence[grant numbers:F2015201033,F2017201069] the foundation of H3C[grant number:2017A20004]。 

主　　题：Deep convolutional neural network Weight initialization Sparse autoencoder Visual saliency model Image recognition 

摘      要：Since of the scale and the various shapes of down in the image,it is difficult for traditional image recognition method to correctly recognize the type of down image and get the required recognition accuracy,even for the Traditional Convolutional Neural Network(TCNN).To deal with the above problems,a Deep Convolutional Neural Network(DCNN)for down image classification is constructed,and a new weight initialization method is proposed.Firstly,the salient regions of a down image were cut from the image using the visual saliency model.Then,these salient regions of the image were used to train a sparse autoencoder and get a collection of convolutional filters,which accord with the statistical characteristics of dataset.At last,a DCNN with Inception module and its variants was constructed.To improve the recognition accuracy,the depth of the network is deepened.The experiment results indicate that the constructed DCNN increases the recognition accuracy by 2.7% compared to TCNN,when recognizing the down in the images.The convergence rate of the proposed DCNN with the new weight initialization method is improved by 25.5% compared to TCNN.