Hand segmentation using Depth thresholding

Question

I want to segment hand from a depth image using depth thresholding. I used this kinect and leap dataset from this link-

http://lttm.dei.unipd.it/downloads/gesture/

I tried these 2 codes, but the output I got is total black image in both the cases. The original .png image is

I selected depth value from 1_depth.bin file in the dataset.

Code 1

I = fopen('D:\dsktop\kinect_leap_dataset\acquisitions\P1\G1\1_depth.bin', 'r');
A = fread(I, 480*640, 'uint8=>uint8');
A = reshape(A, 480, 640);

min_row = min(A);
min_col = min(min_row);

for i = 1:480
    for j = 1:640
        if ((A(i,j) > (min_col + 10)) || (A(i,j) == (min_col + 10)))
           A(i,j) = 1;
       else
           A(i,j) = 0;
        end
    end
end
imshow(A)

Code 2

image = imread('D:\dsktop\kinect_leap_dataset\acquisitions\P1\G1\1_depth.png');
I = fopen('D:\dsktop\kinect_leap_dataset\acquisitions\P1\G1\1_depth.bin', 'r');
A = fread(I, 480*640, 'uint8=>uint8');
A = reshape(A, 480, 640);

min_row = min(A);
min_col = min(min_row);
for i = 1:480
    for j = 1:640
        if ((A(i,j) > (min_col + 10)) || (A(i,j) == (min_col + 10)))
            image(i,j) = 1;
        else
            image(i,j) = 0;
        end
    end
end
imshow(image)

The output I am getting is

Kindly tell what is wrong in this code and why I am not getting any out?

Your `A` is of type `uint8` and has values `{0,1}` to display "white" Matlab expects uint8 value of 255. Try `imshow(A*255);` or `imshow(A,[]);` or `imagesc(A);`. — Shai, Jul 31 '17 at 06:53
It seems like your depth info `A` is stored in "row major" order, as opposed to matlab's "column major". Try read A: `A = fread(I, 480*640, 'uint8=>uint8'); A = reshape(A, 640,480).';` — Shai, Jul 31 '17 at 07:39
when the image is read, MATLAB is showing 480*640 uint8. So I suppose, it is 8bits per pixel. — Prachi, Aug 01 '17 at 06:38
you read it as 8bits per pixel, so this is what Matlab shows. Are you sure this is the right format? — Shai, Aug 01 '17 at 06:41
Variable A (which is a .bin file) is showing 480*640 double. The final "mask" image created according to your code is showing 480*640 logical. Is this wrong for display? — Prachi, Aug 01 '17 at 10:41
`A` is uint8 because you read it as such. The question is is the data in the corresponding '.bin' file is also uint8? what is the file size in bytes? do you have any documentation on how to read this file? — Shai, Aug 01 '17 at 20:09
I checked the read me file of the dataset. The .bin file is "short 16 bit". I haven't checked it before. So I have to use 'int16=>int16' instead of 'uint8=>uint8'. — Prachi, Aug 02 '17 at 05:24
Thanks Shai!! Now the picture is coming proper black and white :) Thanks for bearing with me on this silly question. ;D — Prachi, Aug 02 '17 at 05:27

score 0 · Accepted Answer · answered Jul 31 '17 at 06:57

0

Your code is extremely not vectorize. Here's how to re-write your code in a more vectorize fashion. This is both more efficient and more "readable":

I = fopen('D:\dsktop\kinect_leap_dataset\acquisitions\P1\G1\1_depth.bin', 'r');
A = fread(I, 480*640, 'uint8=>uint8');
A = reshape(A, 480, 640);

min_ = min(A(:));  % minimal value across rows and columns
mask = A>=(min_+10);  % no need for loop, vectorize code.
imshow(mask, []);

answered Jul 31 '17 at 06:57

Shai

111,146
38
238
371

Thanks Shai!! Now I am getting some output other than the black window. :) But I am getting some weird output now which has some black portion but the white portion is not that white, it is somewhat off white. Can you tell why? I have changed the threshold value several times but never got proper white color. – Prachi Jul 31 '17 at 07:26
@Prachi please add screen shots to your question to illustrate the problem. – Shai Jul 31 '17 at 07:31
I have attached the output image above. – Prachi Jul 31 '17 at 07:35

Hand segmentation using Depth thresholding

1 Answers1