基于聚类的超声射频图像阴影双线性补偿方法  

Bilinear Shadow Compensation Method for Ultrasonic Radio Frequency Image Based on Clustering

作　　者：罗扬[1] 袁艺标[2] LUO Yang;YUAN Yi-biao(School of Software Engineering,Jinling Institute of Technology,Jiangsu Nanjing 211169,China;Nanjing Medical University,Jiangsu Nanjing 211166,China)

机构地区：[1]金陵科技学院软件工程学院,江苏南京211169 [2]南京医科大学,江苏南京211166

出　　处：《计算机仿真》2021年第10期209-212,236,共5页Computer Simulation

摘　　要：待识别目标个体若处于图像阴影区域,其轮廓特征就会出现模糊或者失真丢失的情况。传统超声射频图像阴影补偿方法难以提高原图像的清晰度,且阴影补偿后的图像亮度偏低。为此提出基于聚类的超声射频图像阴影双线性补偿方法。采用固定阈值法将图像分成两组,计算两组图像方差的最大值,确定阈值。采用模糊聚类算法将所有辨识对象划分为某一类属性,获取新聚类中心和起初聚类中心之间的距离差,设置对应门限分割条件完成图像分割。利用加权估计位置插值点灰度值计算双线性插值,最后引入RGB色彩空间抑制蓝光分量,将其转换至HIS色彩空间内,完成图像色度、亮度以及饱和度的双线性补偿。实验证明,所提方法补偿后图像阴影区域比原始图像更加清晰,且非阴影区域图像的特征也具有理想的亮度,实验结果证明了所提方法的应用鲁棒性较强。A bilinear shadow compensation method based on clustering was presented for improving the image brightness after shadow compensation.According to the fixed threshold method,the image was divided into two groups to calculate the maximum variance of the two groups of images,getting the threshold.A Fuzzy clustering algorithm was introduced to divide all the identified objects into a certain class of attributes,thus obtaining the distance difference between the new clustering center and the original clustering center.The corresponding threshold segmentation condition was set to segment the image.The gray value of the weighted estimated position interpolation point was applied to calculate the bilinear interpolation.RGB color space was also used to suppress the blue light component and convert it into his color space to achieve the bilinear compensation of image chromaticity,brightness and saturation.The experimental results show that the method can improve the clarity of both shadow and non-shadow regions,showing that the proposed method has excellent application robustness.

关 键 词：模糊聚类分析 最大类间方差阈值分割法 双线性插值 图像补偿 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]