一种改进的白平衡算法在腹腔镜图像增强中的应用研究  

Application of an Improved White Balance Algorithm in Laparoscopic Image Enhancement

作　　者：余同阳 胡多传 诸志龙 YU Tongyang;HU Duochuan;ZHU Zhilong(School of Electrical Engineering,Anhui Polytechnic University,Wuhu 241000,China;Management Department,Wuxi Defan Medical Technology Co.,Ltd,Wuxi 214000,China)

机构地区：[1]安徽工程大学电气工程学院,安徽芜湖241000 [2]无锡德凡医疗科技有限公司管理部,江苏无锡214000

出　　处：《安徽工程大学学报》2024年第3期44-50,共7页Journal of Anhui Polytechnic University

基　　金：国家级大学生创新创业训练计划项目(JX22021023)。

摘　　要：腹腔镜光源的色温波动导致腹腔镜图像偏离真实的色彩,这可能会造成医生的视觉疲劳并且影响医生对病态组织的检测和诊断。为了解决这个问题,提出了基于Lab模型的最优化自动白平衡算法,在动态阈值算法的基础上,利用白点恢复的程度动态调整白点选择区域,使白点恢复效果达到最佳。本模型相较于其他模型,在白点的检测和筛选过程中可以精准且连续地识别近白色区域,因此选取的参考白点更加合适。由于在对图像进行白平衡时,白点检测参数的改变会影响近白色区域的大小,为了得到效果最好的参数,提出一种反馈调整法,以实现各种偏色图像的最优化白平衡。另外还提出一种新的白平衡评价方式,称为白点评价法,适用于评价腹腔镜图像处理的效果。The color temperature fluctuations of the laparoscopic light source cause the laparoscopic image to deviate from the true color,which may cause visual fatigue for doctors and affect their detection and diagnosis of diseased tissues.To solve this problem,an optimized automatic white balance algorithm based on the Lab model is proposed.On the basis of the dynamic threshold algorithm,the degree of white point recovery is used to dynamically adjust the white point selection area,achieving the best white point recovery effect.Compared to other models,this model can accurately and continuously identify near white areas in the process of white point detection and screening,so the selected reference white points are more suitable.The first main task to propose a feedback adjustment method to achieve optimal white balance for various color biased images,as changes in white spot detection parameters can affect the size of the near white area when performing white balance on images.The second task is to propose a new white balance evaluation method,i.e.the white dot evaluation method,which is suitable for evaluating the effect of laparoscopic image processing.

关 键 词：腹腔镜图像 自动白平衡 Lab颜色空间 反馈调整 白点评价 

分 类 号：U471.15[机械工程—车辆工程]