\r冀 中¹，赵可心¹，张锁平²，李明兵\r²\r
\r
AuthorsHTML:\r冀 中¹，赵可心¹，张锁平²，李明兵\r²\r
\r
AuthorsListE:\rJi Zhong¹，Zhao Kexin¹，Zhang Suoping²，Li Mingbing\r²\r
\r
AuthorsHTMLE:\rJi Zhong¹，Zhao Kexin¹，Zhang Suoping²，Li Mingbing\r²\r
\r
Unit:\r\r1. 天津大学电气自动化与信息工程学院，天津 300072；\r
\r\r2. 国家海洋技术中心，天津 300072\r
\r
\r
Unit_EngLish:\r1. School of Electrical and Information Engineering，Tianjin University，Tianjin 300072，China；
2. National Ocean Technology Center，Tianjin 300072，China\r
\r
Abstract_Chinese:\r有效地识别水下各种鱼类目标具有重要的实际意义和理论价值．鱼类生存环境复杂，由于海洋的极端条件，水下鱼类图像的分辨率低，且图像类间相似度高、类内差异性大，并受光照、角度、姿态等的影响较大，这些因素使得鱼类识别成为一项具有挑战的任务．针对这些难点，提出了一个能够有效进行细粒度鱼类图像分类的深度学习模型．该模型包含空间变换网络和双线性网络两部分，首先利用空间变换网络作为注意力机制，去除图像背景中复杂的干扰信息，选择图像中感兴趣的目标区域，简化后续分类；双线性网络通过融合两个深度网络的特征图提取图像的双线性特征，使得对目标中具有判别性的特定位置有较强的响应，从而识别种类，该模型可以进行端到端的训练．在公开的F4K 数据集上，该模型取得了最好的性能，识别正确率为99.36%，较现有最好算法DeepFish 提高0.56%，此外，发布了一个包含100 类共6 358 张图片的新的鱼类图像数据集Fish100，该模型在Fish100 数据集上的识别正确率高出BCNN算法0.98%．多个数据集上的实验验证了模型的有效性与先进性．\r
\r
Abstract_English:\rEffective classification of various fish species under water has great practical and theoretical significance. Due to the extreme conditions of the ocean，underwater images have very low resolution. Since the living environment is highly complex，fish images have properties of high inter-class similarity，large intra-class variety，and are greatly affected by light，angle，posture etc. These factors make fish classification a challenging task. To cope with these challenges，a deep fine-grained fish imageclassification model is proposed. It consists of a spatial transformer network and a bilinear network. Specifically，the spatial transformer network aims at removing the complex background as an attention mechanism and selecting the region of interest in the image. The bilinear network extracts the bilinear features of the image by fusing the feature maps of two deep networks，so that it responds to the discriminative part of the target. The model can be trained in an end-to-end way. The model achieves its best performance on the public F4K dataset. The recognition accuracy was 99.36%，which was 0.56% higher than the DeepFish algorithm. In addition，a new dataset called Fish100，containing 100 categories of 6 358 images，was released. Accuracy of the model is 0.98% higher than that of the bilinear convolutional neural network(BCNN)model. Experiments on several datasets verified the effectiveness and superiority of the proposed algorithm.\r
\r
Keyword_Chinese:鱼类分类；细粒度分类；空间变换；双线性网络\r

Keywords_English:fish classification；fine-grained classification；spatial transformation；bilinear network\r


PDF全文下载地址:http://xbzrb.tju.edu.cn/#/digest?ArticleID=6226