How can I color an image?
古いコメントを表示
Hello, I'm trying to segment an image using the DBScan algorithm, but the result is black and white. How can I convert the result into a colored image? I think, I need to use the reshape function, but I couldn't figure it out. I have little image processing knowledge. Thanks for the help.
rgbImage = imread('1.jpg');
imshow(rgbImage);
rChannel = rgbImage(:, :, 1);
gChannel = rgbImage(:, :, 2);
bChannel = rgbImage(:, :, 3);
Img = double([rChannel(:), gChannel(:), bChannel(:)]);
% m = Idx, n = Indicator for core points -> (X, Epsilon, MinPTS, Distance)
[Idx, n] = dbscan(Img, 0.3, 50,'Distance','chebychev');
imshow(Idx);
採用された回答
その他の回答 (1 件)
Ergin Sezgin
2023 年 3 月 28 日
1 投票
Hello,
Reshape function only changes the size of an array but keeps the number of elements same. For a color image in RGB color space, you need to constitute three color channels, same sized matrices with your grayscale image. The problem is that when you convert your color image into a grayscale image, the color information is permanently lost and it is not possible to automatically recolorize. You can manually colorize it with a photo editing software but even in that case it will not be the same.
However, in your case the output doesn't seem to be an image. Instead it's an index array, representing which cluster or segment the pixels in your image belong to.
2 件のコメント
Image Analyst
2023 年 3 月 28 日
Like I said in my answer below, kmeans, and probably dbscan, is not a good color segmentation method. Maybe you should try rgb2ind() if you want something simple.
カテゴリ
ヘルプ センター および File Exchange で Image Segmentation についてさらに検索
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!

