查询结果:   王刚,宫元九.基于深度学习的锂电池褶皱检测方法的研究[J].计算机应用与软件,2019,36(1):216 - 219,231.
中文标题
基于深度学习的锂电池褶皱检测方法的研究
发表栏目
人工智能与识别
摘要点击数
888
英文标题
LITHIUM BATTERY WRINKLE DETECTION METHOD BASED ON DEEP LEARNING
作 者
王刚 宫元九 Wang Gang Gong Yuanjiu
作者单位
辽宁大学信息学院 辽宁 沈阳 110036     
英文单位
College of Information, Liaoning University, Shenyang 110036, Liaoning, China     
关键词
深度学习 CNN 锂电池褶皱检测
Keywords
Deep learning CNN Lithium battery wrinkle detection
基金项目
作者资料
王刚,硕士生,主研领域:数字图像处理,深度学习。宫元九,教授。 。
文章摘要
为了解决锂电池褶皱检测问题,提出基于深度学习的褶皱检测方法。CNN是在实际应用中最成功的深度神经网络,能够很好地实现分类。提出CNN来解决锂电池褶皱检测的方法:收集锂电池X光照片;人工把锂电池X光照片分为无褶皱和有褶皱两种,并标注;将数据集放入构建的CNN模型中训练学习。在数据集充足的情况下,通过大量实验表明:该方法的准确率能够达到99%,相对于原始的凭借经验、人工观察检测的方法有很大提升。
Abstract
In order to solve the problem of lithium battery wrinkle detection, we proposed a wrinkle detection method based on deep learning. CNN is the most successful deep neural network in practical applications. It can achieve good classification. CNN was put forward to solve the problem of lithium battery wrinkle detection. We collected the X-ray photo of the lithium battery. The X-ray photo of the lithium battery was manually divided into two types: no wrinkles and wrinkles, and we marked them. The data set was put into the constructed CNN model for training and learning. With sufficient data sets, a large number of experiments have shown that the accuracy of the CNN-based lithium battery wrinkle detection method can reach 99%. Compared with the original experience, and manual observation and detection methods, it has been a big improvement.
下载PDF全文