基于块匹配的全景图像拼接

function [MStitch, im] = Fun_StitchRGB(im2, W_box, H_box, bdown, MStitch) W_min = MStitch.W_min; W_max = MStitch.W_max; H_min = MStitch.H_min; minval = MStitch.minval; im1 = MStitch.imrgb1; [Fheight, Fwidth] = size(im2); im = MStitch.imrgb1; im2 = MStitch.imrgb2; [Pheight, Pwidth, depth] = size(im); sz2 = size(im2); [Pheight2, Pwidth2, depth2] = size(im2); w = 0; hw = waitbar(0, '图像拼接进度：', 'Name', '图像拼接……'); if bdown x2 = 1; for x1 = Pwidth-W_box : Pwidth y2 = 1; for y1 = Pheight-H_box+1 : Pheight /* 开发不易，整理也不易，如需要详细的说明文档和程序，以及完整的数据集，训练好的模型，或者进一步开发， 可加作者新联系方式咨询，WX：Q3101759565，QQ：3101759565 */ end x2 = x2 + 1; rt = 0.5*(x1 - Pwidth + W_box)/W_box; waitbar(rt, hw, sprintf('图像拼接进度：%i%%', round(rt*100))); end rt0 = rt; for y1 = 1 : H_box for x3 = x2 : Pwidth2 [x1, y1] = CheckRC(x1, y1, im1); [x3, y1] = CheckRC(x3, y1, im2); im(y1, Pwidth+x3-x2+1, :) = im2(y1, x3, :); end rt = rt0 + 0.5*(y1 - 1)/H_box; waitbar(rt, hw, sprintf('图像拼接进度：%i%%', round(rt*100))); end else x2 = 1; for x1 = Pwidth-W_box : Pwidth y2 = 1; for y1 = Fheight-H_box+1 : Fheight [x1, y1] = CheckRC(x1, y1, im1); [x2, y2] = CheckRC(x2, y2, im2); w = x2/W_box; im(y1, x1, 1) = im1(y1, x1, 1)*(1.0-w) + im2(y2, x2, 1)*w; im(y1, x1, 2) = im1(y1, x1, 2)*(1.0-w) + im2(y2, x2, 2)*w; im(y1, x1, 3) = im1(y1, x1, 3)*(1.0-w) + im2(y2, x2, 3)*w; y2 = y2 + 1; end x2 = x2 + 1; rt = 0.5*(x1 - Pwidth + W_box)/W_box; waitbar(rt, hw, sprintf('图像拼接进度：%i%%', round(rt*100))); end rt0 = rt; for y1 = Fheight-H_box+1 : Fheight for x3 = x2 : Fwidth [x1, y1] = CheckRC(x1, y1, im1); [x3, y1] = CheckRC(x3, y1, im2); im(y1, Pwidth+x3-x2+1, :) = im2(y1, x3, :); end rt = rt0 + 0.5*(y1 - 1)/H_box; waitbar(rt, hw, sprintf('图像拼接进度：%i%%', round(rt*100))); end end MStitch.imrgb1 = im; [Pheight, Pwidth, depth] = size(im); MStitch.Pwidth = Pwidth; MStitch.Pheight = Pheight; rt = 1; waitbar(rt, hw, sprintf('图像拼接进度：%i%%', round(rt*100))); delete(hw);