基于深度学习的疟疾检测  

Malaria detection based on in-depth learning

作　　者：刘银萍[1] 尹明[2] 陈平[2] 曾奕秋 LIU Yinping;YIN Ming;CHEN Ping;ZENG Yiqiu(Experiment Teaching Department,Guangdong University of Technology,Guangzhou 510006,China;School of Automation,Guangdong University of Technology,Guangzhou 510006,China)

机构地区：[1]广东工业大学实验教学部,广东广州510006 [2]广东工业大学自动化学院,广东广州510006

出　　处：《实验技术与管理》2020年第2期67-71,共5页Experimental Technology and Management

基　　金：国家自然科学基金面上项目(61876042).

摘　　要：针对疟疾检测方法中的模型存在训练时间过长,权重参数冗余等问题,用疟疾数据集从头开始训练,更改输入图像的大小,直接对ResNet-50网络的深度和宽度进行缩减,研究采用深度学习技术快速、准确地检测疟疾。该方法缩短了模型训练时间,提高了疟疾分类精确度,缩小了模型权重参数大小。In view of the problems of long training time and redundant weight parameters in the model of malaria detection method,we use the malaria data set from the beginning for training,change the size of input image,directly reduce the depth and width of resnet-50 network,and study the rapid and accurate malaria detection method by using the in-depth learning technology.This method shortens the training time of the model,improves the accuracy of malaria classification and reduces the weight parameters of the model.

关 键 词：疟疾检测 深度学习 从头训练 

分 类 号：R531.3[医药卫生—内科学] TP391.41[医药卫生—临床医学]