MATLAB Image Processing
$30-250 USD
Dibayar semasa penghantaran
Extend an existing MATLAB script that performs k-means cluster analysis for image segmentation to:
1) Convert centroid colors from LAB a*,b* color to RGB; assuming mid-range luminance;
2) Calculate the variance for each cluster
3) For each cluster, plot the color of each pixel of the input image onto an xy plot with axes a* and b*. That is, for the "gold" partition, take all of the pixels in the input image that are labeled "gold" and plot their color on a graph with axes a* and b*.
The attached MATLAB script performs k-means cluster analysis to segment images. This is nearly verbatim from
[url removed, login to view]
and it relies on the Statistics Toolbox.
An example image is included.
This is a work-for-hire and all intellectual property right are assigned to me.
For someone who knows that they are doing, this should be a couple hours work, max. Difficulties I've run into:
1) Centroid is returned as a*,b* from LAB color space. However, the values of a*,b* are out of the expected range of about [-110,110]. Therefore, conversion back to RGB doesn't work.
2) Logical indexing of matrices is new to me (in MATLAB). When I try to calculate the variance, I have mismatched argument sizes.
ID Projek: #4668698