is Arnold transform for color image and for grey/binary image different?

Question

anika hossain 2015-10-5

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/246806-is-arnold-transform-for-color-image-and-for-grey-binary-image-different

评论： anika hossain 2015-10-6

i need to scramble a rgb color image. now i want to know is Arnold transform for color image and for grey/binary image different? it will be great if any one can refer the appropriate code from file exchange

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Walter Roberson 2015-10-5

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/246806-is-arnold-transform-for-color-image-and-for-grey-binary-image-different#answer_194707

The 3D version is different; see http://www.mathworks.com/matlabcentral/answers/167213-3-dimensioal-arnold-cat-map

However you could reshape() your color image to be rows by (cols * 3), and then pad the result out to square and apply the arnold transform to that.

3 个评论
显示 1更早的评论隐藏 1更早的评论

Walter Roberson 2015-10-5

There is no such thing as a 2D rgb color image in MATLAB. In MATLAB, 2D images can be pseudocolor ("color mapped"), or you can have 3D arrays which are RGB "true color". If you have a 2D array of data that represents color then you should just do a normal arnold transform on it and apply the colormap any time you want to display the result. If you have a 3D truecolor array, then you could divide it into channels and apply the arnold map to each, or you could reshape to have 3 times as many columns and apply the arnold map to that, or you could apply the 3D cat map.

anika hossain 2015-10-6

okay,i first divided the image in red,green, blue channel so i have 2D array now and then apply arnold cat map in individual three channels with same iteration. if the key is correct the reverse should be found identical to the real one right? and it is identical when i don't do any other things between the arnold apply and inverse arnold.but when i do some processing (embed watermark which only modify 1/4 of the image's pixel value with less then 1) and then apply inverse the image it too much destroyed to be accepted. why? and should i try 3D cat map instead of that?

请先登录，再进行评论。