文章摘要
王基实;杨珺涵;张世祺;高诗岳;高爽;杨关林;张哲.基于GWO-BP神经网络在舌诊图像颜色校正中的应用[J].中医药信息,2024,41(5):31-36
基于GWO-BP神经网络在舌诊图像颜色校正中的应用
Application of GWO-BP Neural Network in Color Correction of Tongue Diagnosis Images
投稿时间:2023-09-27  录用日期:2024-01-08
DOI:10.19656/j.cnki.1002-2406.20240506
中文关键词: 舌诊图像  颜色校正  灰狼优化算法  BP神经网络
英文关键词: Tongue diagnosis images  Color calibration  Grey Wolf Optimization algorithm  BP neural network
基金项目:国家中医药管理局首批青年岐黄学者支持项目(文件号:国中医药人教函[2020]218号)
作者单位
王基实;杨珺涵;张世祺;高诗岳;高爽;杨关林;张哲  
摘要点击次数: 41
全文下载次数: 43
中文摘要:
      目的:基于灰狼优化算法(GWO)优化传统BP神经网络的舌诊图像颜色校正算法,解决移动舌诊APP存在的拍摄环境具有局限性、手机设备的依赖性及基于手机平台的颜色校正方法所使用的传统算法效果不佳等问题。方法:以24色标准色卡为标准,分别采集室内白炽灯不同光照强度场景与室外不同时刻阴天、晴天场景下的舌诊图像;同时选择灰度世界算法以及传统BP神经网络算法与本文算法进行对比,分别用上述三种算法对采集图像进行颜色校正,并对校正结果进行主观与客观颜色评价进行对比分析。结果:相比于灰度世界算法及BP神经网络算法,GWO-BP神经网络算法的颜色校正效果明显提高。结论:GWO-BP算法可有效地对手机拍摄的舌诊图像进行颜色校正,从而提高色值精确度。
英文摘要:
      Objective: Based on the Grey Wolf Optimization(GWO) algorithm, to optimize the tongue diagnosis image color correction algorithm of the traditional BP neural network, and to solve the problems such as the limitation of shooting environment of mobile tongue diagnosis APP and the dependence of mobile phone devices and the poor effect of traditional algorithm used by color correction method based on mobile phone platform. Methods: Using 24-color standard color card as the standard, the tongue diagnosis images of indoor incandescent lamp under different light intensity and outdoor cloudy and sunny scenes at different times were collected. At the same time, grayscale world algorithm and traditional BP neural network algorithm were selected to compare with the proposed algorithm of this paper. The above three algorithms were used to correct the color of the acquired image, and the correction results were compared with objective and subjective color evaluation. Results: Compared with gray-scale world algorithm and BP neural network algorithm, the color correction effect of GWO-BP neural network algorithm was significantly improved. Conclusion: The GWOBP algorithm can effectively correct the color of tongue diagnosis images taken by mobile phones, so as to improve the accuracy of color values.
查看全文   查看/发表评论  下载PDF阅读器
关闭
微信公众号
分享按钮