瞳孔动态监测系统的研究与开发.docxVIP

瞳孔动态监测系统的研究与开发摘 要本文以红外线条件下录制的瞳孔视频为处理对象，对瞳孔视频进行预处理，得到用特殊颜色标识的瞳孔视频为输出。通过对视频文件的逐帧采集，对采集的帧图像，进行图像处理得到瞳孔的面积大小并用特殊颜色标识瞳孔，然后进行帧合成为视频输出文件。本文采用对采集的帧图像进行腐蚀、膨胀、开运算、二值化等图像预处理的方法，并用区域生长算法得到瞳孔的区域，从而计算出瞳孔区域的相对像素数，可以计算出瞳孔面积大小，并绘制瞳孔面积变化曲线，然后将处理后的图像帧合成为视频输出文件，从而实现瞳孔的动态分析。本文对帧的处理是基于二值化的粗定位和区域生长方法相结合的方法，能有效检测因病变而呈不规则圆形的瞳孔，简单而且快速。首先使用形态学开运算进行预处理，去除睫毛和亮斑的影响且保持原有瞳孔大小，同时生成缩略图；然后生成二值阈值化图像，进行水平和垂直投影，找到极值点，将其作为种子点；最后用区域生长算法得到精确的瞳孔区域。基于以上的处理方法，采用OpenCV、VC++编程工具，编程实现了瞳孔动态监测系统的软件部分。关键词：二值化；开运算；区域生长法；瞳孔定位；C++AbstractBased on the infrared body under the condition of recording pupil video as treatment object, the pupil video preprocessing, get with the special color identification pupil video for the output. Through the frame by frame of video file collection, the collection of image, image processing of pupil size and special color mark the pupil, then the frame synthesis for video output file.This article adopts the corrosion, expansion, with the frame of open operation, binary image preprocessing methods, such as using region growing algorithm to get the pupil area, so as to calculate the relative sizes of pupil area, can calculate the pupil size, and draw the pupil area change curve, and then synthesize processed image frames into video output file, so as to realize the dynamic analysis of the pupil. In this paper, the frame is based on the processing of binarization of coarse location and region growing method with the combination of methods, can effectively detect irregular round pupil by lesions, simple and rapid. Preprocessing of the first to use morphological open operation to remove eyelash and bright spots and keep the pupil size, the influence of the generated thumbnail at the same time; Then generate binary threshold images, the horizontal and vertical projection, find the extreme value point, as the seed point; Finally get accurate pupil area using region growing algorithm. Processing method based on the above, the use of OpenCV, vc + + programming tools, programming software part of