基于差分法的图像粒子运动分析  

Movement Analysis of Particles Based on Image Difference

作　　者：朱建新[1] 严建华[1] 马增益[1] 丁经纬[1] 池涌[1] 岑可法[1] 

机构地区：[1]能源洁净利用与环境工程教育部重点实验室,浙江大学热能工程研究所,杭州310027

出　　处：《中国图象图形学报（A辑）》2004年第4期402-407,共6页Journal of Image and Graphics

基　　金：国家自然科学重点资助项目(59836210);国家重点基础研究专项经费资助(2001CB409600)

摘　　要：利用差分法配合灰度拉伸对颗粒运动图像进行了处理 ,直方图分析表明 ,像素在灰度上的分布特性与颗粒运动的剧烈程度具有相关性。差分图像中高亮度像素越多 ,颗粒运动越剧烈。随运动剧烈程度的降低 ,灰度相近的像素 ,其亮度由高向低变化。除去低亮度区域内白噪声的影响后 ,以对数函数为权值 ,在考虑了一定亮度范围内的像素分布后 ,提出了一个用于评价颗粒运动剧烈程度的指数。将其应用于流化床气泡生成图像序列表明 。By using image difference and gray stretch technique, image histogram analysis method was used to investigate the movement of particles in image sequence. The result shows that the movement of particles has close relation with the dispersivity of pixels' greyvalue. The number of pixels with high brightness indicates the fierceness of particles. The brightness of pixels with similar grayscale will decrease with the slowdown of particles. After eliminating the influence of noise and taking consider of the gray level of pixels within certain range, a weight factor based on logarithm function was introduced to evaluate the intensity of particles' movement. Application of this algorithm in image sequence of fluidized bubble verifies the validity of this method.

关 键 词：图像处理 差分法 运动指数 图像序列 直方图分析 流化床 

分 类 号：TQ051.13[化学工程] TP391.41[自动化与计算机技术—计算机应用技术]