How do i retrieve image based on its value in an excel file?

Dear Experiences...

i have an excel file.. in the following directory (D:\image_table).. where in this table there 100 x 100 matrix value look like the following, where the values refer to the degree of similarity among images (0..1), and images are store in folder (D:\images).. images names are (A1...A100).

          A1    A2     A3   A4   A5   A6   .... A100  
  A1     1.0   0.25  0.33 ....
  A2     0.25  1.0 ..................
  A3
  A4
  ..
  ..
  A100

so, i need to retrieve all images names that satisfy condition like similarity>0.50 and also view all images from D:\images folder in category that match images names.. perhaps images name view as table.. thanks

excel file..

[Image Analyst]

Where are the image names stored? Are they the row and column headers? Can you attach the workbook?

[ahmed obaid]

yes, images name are row and column header... this table look like pairwise similarity table.. where image names are A1...A100

Answer 1

[Image Analyst]

Try this:

[rows, columns] = find(similarityMatrix > 0.5);
for k = 1 : length(rows)
    fprintf('%s and %s have a similarity of %f.\n', ...
    filenames(rows(k)), filenames(columns(k)), ...
    similarityMatrix(rows(k), columns(k));
end

[ahmed obaid]

its good till now,

for query image look like the following line:

query=imread(D:\images\A1.png');

according to attached excel file.. the retrieved images are( A3 and A5), then from the given folder D:\images.. i need to view similar images in montage view. how i do that please

[ahmed obaid]

attached excel file

[Image Analyst]

I don't understand the question, and can't find any attachment.

To stitch together images side-by-side:

wideImage = [image1, image2]; % Must have same number of rows.

To stitch together above and below:

tallImage = [image1; image2]; % Must have same number of columns.

[ahmed obaid]

attached file

[ahmed obaid]

please find the attached data with my best regards

[Image Analyst]

Let's say you have a random smattering of values in the array that are more than 0.5, so you have a bunch of image pairs. I'm not sure what you want. Do you want a wide image with the pair of images stitched side-by-side? Or above and below? Or something else? So if you had 5 numbers, you'd have 5 stitched image pairs? Please try to explain better. Attaching a desired output image would also be great.

[ahmed obaid]

Thank you for your time and explanation, when excel file import to matlab either by used readtable or import data, just imagine that first row contain A1 header (row_header), then if i read image using:

image=imread(D:\images\A1.png); 
then by use [~, baseFileNName, ~]= fileparts(image.name);

we can know what is the name of reading image.. then suppose image name is A1, now im interesting just on first row.. where A1 have similarity values to all other images include A1, now we are check values of first row .. when value >=0.5 then result must shown (according to attached file) { A1, A3 and A5} as matched images.. then using path of my folder (D:\images) to view result images (A1, A3 and A5) in montage view or such other method...

perhaps this requirements can be applied... best regards.