Deng Jiehang, He Dongdong, Zhuo Jiahong, Zhao Jian, Xiao Cheng, Kang Xiaodong, Hu Sunlin, Gu Guosheng, Liu Chao
School of Computer Science, Guangdong University of Technology, Guangzhou 510006, China.
School of Computer Science and Software, Zhaoqing University, Zhaoqing 526061 China.
Nan Fang Yi Ke Da Xue Xue Bao. 2020 Feb 29;40(2):183-189. doi: 10.12122/j.issn.1673-4254.2020.02.08.
We propose a deep learning network-based method for recognizing and locating diatom targets with interference by complex background in autopsy.The system consisted of two modules: the preliminary positioning module and the accurate positioning module. In preliminary positioning, ZFNet convolution and pooling were utilized to extract the high-level features, and Regional Proposal Network (RPN) was applied to generate the regions where the diatoms may exist. In accurate positioning, Fast R-CNN was used to modify the position information and identify the types of the diatoms.We compared the proposed method with conventional machine learning methods using a self-built database of images with interference by simple, moderate and complex backgrounds. The conventional methods showed a recognition rate of diatoms against partial background interference of about 60%, and failed to recognize or locate the diatom objects in the datasets with complex background interference. The deep learning network-based method effectively recognized and located the diatom targets against complex background interference with an average recognition rate reaching 85%.The proposed method can be applied for recognition and location of diatom targets against complex background interference in autopsy.
我们提出了一种基于深度学习网络的方法，用于在尸检中识别和定位受复杂背景干扰的硅藻目标。该系统由两个模块组成：初步定位模块和精确定位模块。在初步定位中，利用ZFNet卷积和池化来提取高级特征，并应用区域提议网络（RPN）生成硅藻可能存在的区域。在精确定位中，使用快速R-CNN来修改位置信息并识别硅藻的类型。我们使用自建的具有简单、中等和复杂背景干扰的图像数据库，将所提出的方法与传统机器学习方法进行了比较。传统方法对部分背景干扰的硅藻识别率约为60%，在具有复杂背景干扰的数据集中无法识别或定位硅藻目标。基于深度学习网络的方法有效地识别和定位了受复杂背景干扰的硅藻目标，平均识别率达到85%。所提出的方法可应用于尸检中受复杂背景干扰的硅藻目标的识别和定位。
