文件列表:

color_histeq.m (1474, 2010-03-18)
dark_channel.asv (1248, 2010-12-18)
dark_channel.m (1346, 2010-12-18)
hazefree.m (582, 2010-03-18)
matLap.asv (3652, 2010-12-27)
matLap.m (3653, 2010-03-18)
refine_t.m (477, 2010-03-18)
trans.asv (2557, 2010-12-23)
trans.m (2557, 2010-12-23)

% Haze Free Process README file % Note that this may be a memory intensive process, so the user may want to % perform each step individually to make sure of completion. % Step 1, load input image with double precision, normalized to 1 I = double(imread('canyon-input.png'))./255; % Step 2, Compute coarse transmission map and airlight estimation w = 0.95; % Preserve 5% of haze [tc A] = trans(I, w); % Step 3, Compute matting laplacian. % Note that this may take several minutes, depending on image size. % On an AMD Neo X2 with 4GB RAM, a 600x450 color image takes ~88 seconds L = matLap(I); % Step 4, Compute refined transmission map t = refine_t(tc, L); % Step 5, Compute final, haze-free image J = hazefree(I, t, A); % Step 6, OPTIONAL: Image may be too dark. Color histogram processing might % be helpful. User may want to try the supplied color_histeq function. [JI JH JA] = color_histeq(J); % Note that this gives 3 different types of results: imadjust, histeq, and % adapthisteq. All are performed on luminosity.

近期下载者：

相关文件：

评论：[我要评论] [举报此文件]

收藏者：