图像中烟雾检测

tic; clear all; close all; clc; A = aviinfo('smoke.avi'); LEN = A.NumFrames;%得到视频的帧数 for x = 1 : LEN pic = imread(strcat('smoke', int2str(x), '.jpg'));%读入一张图片 pic1 = pic;%pic1中存放的是处理过后的彩色图像 pic_hsv = rgb2hsv(pic); s = size(pic1); bw_pic = zeros(s(1), s(2)); for i = 1 : s(1) for j = 1 : s(2) if pic_hsv(i, j, 1) > 0.35 && pic_hsv(i, j, 1) < 0.65 && ... pic_hsv(i, j, 2) > 0.06 && pic_hsv(i, j, 2) < 0.3 && ... pic_hsv(i, j, 3) > 0.5 && pic_hsv(i, j, 3) <= 1; pic1(i, j, 1) = 255; pic1(i, j, 2 : 3) = 0; end if pic1(i, j, 1) == 255 && pic1(i, j, 2) == 0 bw_pic(i, j) = 1; end end end cov_bw_pic = 1 - bw_pic; fill_pic = imfill(cov_bw_pic, 'holes'); fill_pic1 = 1 - fill_pic; B_pic = bwboundaries(fill_pic1);%得到的是一个P×1的单元 ss = size(B_pic);%ss的值就是连通分量的数量 ss1 = size(B_pic{1, 1});%ss1是一个大小为m×2的矩阵 max = ss1(1); for i = 2 : ss(1)%找到最大的连通区域的边界总个数 ss2 = size(B_pic{i, 1}); if ss2(1)> max max = ss2(1); end end %⑤…………………………将小于最大连通区域的连通分量边界赋零 for k = 1 : ss(1)%将小于最大连通区域的连通分量边界赋零， %其作用就是将其余小的连通分量与边界隔开 Mid = B_pic{k, 1}; ss3 = size(Mid); if ss3(1) < max for i = 1 : ss3(1) fill_pic1(Mid(i, 1), Mid(i, 2)) = 0; end end end %反转填充 cov_fill_pic1 = 1 - fill_pic1; A_fill_pic = imfill(cov_fill_pic1, 'holes');%进行内部填充，再反变换回来，即得到所要求的区域 A_fill_pic1 = 1 - A_fill_pic; B_pic1 = bwboundaries(A_fill_pic1); sss = size(B_pic1);%sss的值就是连通分量的数量 for k = 1 : sss(1)%该循环作用是找到最大区域的边界，然后在原图像的对应位置上赋红 Mid = B_pic1{k};%Mid是一个m×2的矩阵，m就是边界点的数量 %ss4(1)表示边界点的数量，Mid(i, 1)和Mid(i, 2)表示边界点的坐标 ss4 = size(Mid); if ss4(1) == max for i = 1 : ss4(1) pic(Mid(i, 1), Mid(i, 2), 1) = 255; pic(Mid(i, 1), Mid(i, 2), 2 : 3) = 0; end end end imwrite(pic, strcat('lou_smoking', int2str(x), '.jpg'), 'jpg'); end aviobj = avifile('lou_smoking.avi');%将处理后的图片连成视频文件 aviobj.fps = 25;%初始化视频文件的属性 aviobj.Quality = 100; aviobj.compression = 'None'; for i = 1 : LEN frame = imread(strcat('lou_smoking', int2str(i), '.jpg')); aviobj = addframe(aviobj, uint8(frame)); end aviobj=close(aviobj);%关闭aviobj的对象 toc;