MATLAB Answers


I have divided image into 30 blocks and obtained histogram of each block using LBP algorithm, i do not know how to store them separately and merge all histogram to get the single histogram?I want code for merging and saving each histogram of block.

Asked by radha
on 29 Dec 2013
Latest activity Answered by Nikolay S. on 16 Mar 2015
If I create array of Histogram then it gives me error that Subscripted assignment dimension mismatch. so please help me.


If you want then I can send you my .m file. Sorry I also Want to Know how to create .mat file because I am using Matlab R2011-a Version and in that i can not find any function to create .mat file. Thank you for your response. Please help me.
Please just help me to concatenate all histogram of an image into single histogram., and how to create .mat file in R2011a version of matalb??
I do not have any idea about merging/concatenation of histograms. so your help is of great importance to me. I am doing project on this so I expect that you reply as son as possible.

Sign in to comment.

2 Answers

Answer by Image Analyst
on 7 Jan 2014
 Accepted Answer

Create a mat file with the save() function. You can attach your actual image with the paper clip icon if your code can only work with that kind of image. To add histograms, do this
sumHist = hist1 + hist2; % Must have same number of bins. hist1 can be sumHist if you want.


I have 100 images in the folder. so how I loop through it so that I can save all 100 histograms in .mat file?? Thank you
Can you please also help me for implementing the spatial pyramid matching kernel?? I mean I don't have any idea about its implantation that how I should proceed?? So please help me for these two implementation i.e. spatial pyramid matching and one I already asked about .mat file in above comment.
Thanks in advance. You really helped me. Now just I have these two things need to implement. I hope you also help me for that as well.

Sign in to comment.

Answer by Nikolay S. on 16 Mar 2015

Hi there. I'd suggest to have the histograms concatenated via Matlab CAT command. masterHist=cat(1, hist1, hist2,...hist30);
I believe a similar scheme is proposed for face detection in "Face Description with Local Binary Patterns: Application to Face Recognition",
Best regards, Nikolay


Sign in to comment.