Bilinear Interpolation Adaptive Histogram Equalization matlab代码实现

以下是Bilinear Interpolation Adaptive Histogram Equalization的Matlab代码实现：

function [outImg] = bilinearAdapthisteq(inImg, numTiles, clipLimit)

% Bilinear Interpolation Adaptive Histogram Equalization
% inImg: input image
% numTiles: number of tiles (default: 8x8)
% clipLimit: clipping limit (default: 0.01)
% outImg: output image

if nargin < 2
    numTiles = [8 8];
end

if nargin < 3
    clipLimit = 0.01;
end

inImg = double(inImg);

[h, w] = size(inImg);

tileH = floor(h / numTiles(1));
tileW = floor(w / numTiles(2));

outImg = zeros(h, w);

for i = 1:numTiles(1)
    rowStart = (i - 1) * tileH + 1;
    rowEnd = min(rowStart + tileH - 1, h);
    rowSize = rowEnd - rowStart + 1;
    
    for j = 1:numTiles(2)
        colStart = (j - 1) * tileW + 1;
        colEnd = min(colStart + tileW - 1, w);
        colSize = colEnd - colStart + 1;
        
        tile = inImg(rowStart:rowEnd, colStart:colEnd);
        
        % histogram equalization
        tileHist = histcounts(tile(:), 256);
        tileHistCum = cumsum(tileHist) / sum(tileHist);
        tileHistNew = max(0, floor(tileHistCum(tile(:)) * 255));
        
        % clip histogram
        tileHistNewClip = tileHistNew;
        tileHistNewClip(tileHistNewClip > clipLimit * rowSize * colSize) = clipLimit * rowSize * colSize;
        
        % bilinear interpolation
        tileBilinear = interp2((0:255)', repmat([0 255], 256, 1), [tileHistNewClip; tileHistNewClip], tile, 'linear');
        tileBilinear(isnan(tileBilinear)) = 0;
        
        outImg(rowStart:rowEnd, colStart:colEnd) = tileBilinear;
    end
end

outImg = uint8(outImg);

该函数的输入为原始图像、瓦片数量和剪切限制。输出为进行Bilinear Interpolation Adaptive Histogram Equalization后的图像。在实现中，将原始图像分成多个瓦片，对每个瓦片进行直方图均衡化。然后，通过双线性插值将直方图映射应用于每个像素。最后，将所有瓦片重新组合为一幅图像。