how can extract the vein only from the finger vein image?

i want to extract the vein portion only from this image. what method is used to adopt for this? please send important steps for it.

Answer 1

[Image Analyst]

Possibly a Hessian or Frangi filter - often used in angiography. But check the literature http://www.visionbib.com/bibliography/contentsmedical.html#Medical%20Applications,%20CAT,%20MRI,%20Ultrasound,%20Heart%20Models,%20Brain%20Models

and http://www.visionbib.com/bibliography/contentspeople.html#Face%20Recognition,%20Detection,%20Tracking,%20Gesture%20Recognition,%20Fingerprints,%20Biometrics

[ROSE]

i cant acccess this site properly. so please show the codes only for vein extract

[Image Analyst]

I don't have the codes. And there is no finger vein finding routine in MATLAB so you'll have to build it up from lower level routines. Or better yet, just buy it from someone who has already spent years perfecting the code. Contact the authors of papers and offer to buy their code.

It's weird that your ISP is blocking that site. I don't see any reason why you should not be able to get to that particular site. Try this one for the medical literature: https://www.ncbi.nlm.nih.gov/pubmed/ Hopefully you can get to that one.

Answer 2

[ROSE]

i want to extract the vein from this image. for that i use several steps, first i re-size the image into a 150*100 then divide this image into 2 equal half,now my image is in two blocks,upper and lower. now i want to create a mask and divide the mask.then set the upper part of the mask to -1 and lower part of the mask to 1.Then filter the image using imfilter function.use 'replicate'.again divide the filtered image into upper/lower half.

for this steps, i successfully do some beginning portion shown below. please complete the remaining steps using mat codes. clc; i=imread('C:\Us\new1.jpg'); figure(1); imshow(i); i=rgb2gray(i); img=imresize(i,[150 100]); n=fix(size(img,1)/2); A=img(1:n,:,:); %% UPPER PART B=img(n+1:end,:,:); %% LOWER PART figure(2); imshow(A); figure(3); imshow(B); mask=zeros(4,40);

[Image Analyst]

Doesn't make sense to me. Why create a mask? What is it ever used for? What filter did you use in imfilter() and why? "use replicate again" - what do yo umean by "again"? When did you ever even use it the first time?

[ROSE]

i cant access the file you send. i try to extract the vein from the given image. now you please help me to create a mask and divide it equally,upper give the value -1 and lower to 1. very urgent

[Image Analyst]

I did not send a file to you, or even upload a file. The closest thing I did to any of that was to tell you a web site URL.

To make a mask, you can do

[rows, columns, numberOfColorChannels] = yourImage;
mask = ones(rows, columns); % Initialize to all +1
middleRow = floor(rows/2);
mask(1:middleRow, :) = -1; % Set top half to -1.

[ROSE]

i cant understand the first step.why it use? what do you mean by color channels?

i want to create a mask of full zeros with a size 4*40 and divide it into equally and set upper part as -1 and lower part as 1. please send the codes for it....

[Image Analyst]

A mask is a 2D image. Sometimes images are 3D (color) and if you don't ask for that third output of size, you'll get the number of columns as the number of columns times the number of color channels. See this link

I did send you the code for it. Did you see it? If the number of rows in your image is 4, and the number of columns is 40, you'll get a 4-by-40 mask.

Answer 3

[ROSE]

how to find the maximum valu and its idices of an image?

[Image Analyst]

Try this

% Find max value:
maxValue = max(grayImage(:));
% Find all rows and columns where this occurs:
[rows, columns] = find(grayImage == maxValue);

Answer 4

[ROSE]

what is the use of a 'replicate' function using 'imfilt' ?

[Image Analyst]

There is no function by that exact name. There is a rempat. But when you said that you "use 'replicate'.again divide the filtered image into upper/lower half.' in your algorithm, I have no idea why repmat() or some other replication routine would be used to divide an image into upper and lower halves. It's just not needed as far as I can tell because you can just use simple indexing - nothing needs to be replicated or duplicated.

[ROSE]

</matlabcentral/answers/uploaded_files/52737/44.PNG> i want to extract the finger region first.

[ROSE]

i almost complete the program that i already said before. but there is an error showing at line no 28 like this Subscript indices must either be real positive integers or logicals.

Error in trial (line 28) ext_img(ind_up(j):ind_lo(i)+size(filtimg_lo,1), j) = 1; please correct the program........

clc; i=imread('C:\Users\sony\Desktop\lekshmii\new1.jpg');%figure(1);imshow(i); img=rgb2gray(i);figure(2),imshow(i); img=imresize(img,[150 100]); % figure(3),imshow(img),title('resized'); [img_h,img_w]=size(img);

if mod(img_h,2) == 0 half_img_h = (img_h/2 )+ 1; else half_img_h = ceil(img_h/2); end mask=zeros(4,40); %% CREATING A MASK.. [msk_h,msk_w]=size(mask); mask(1:msk_h/2, :) = -1; mask(msk_h/2+1:end , :) = 1;figure(4),imshow(mask);

filtimg = imfilter(img,mask,'replicate');

filtimg_up=filtimg(1:(img_h/2), :); [~,ind_up]=max(filtimg_up);

filtimg_lo=filtimg((img_h/2)+1:end, :); [~,ind_lo]=min(filtimg_lo);

ext_img = zeros(size(img)); for j=1:img_w ext_img(ind_up(j):ind_lo(i)+size(filtimg_lo,1), j) = 1; end figure(7),imshow(ext_img); %Save y-position of finger edges edges = zeros(2,img_w); edges(1,:) = ind_up; edges(2,:) = round(ind_lo + size(filtimg_lo,1));

[Image Analyst]

See this: http://www.mathworks.com/matlabcentral/answers/13205-tutorial-how-to-format-your-question-with-markup

and this FAQ entry: http://matlab.wikia.com/wiki/FAQ#How_do_I_fix_the_error_.22Subscript_indices_must_either_be_real_positive_integers_or_logicals..22.3F

and possibly this link: http://blogs.mathworks.com/videos/2012/07/03/debugging-in-matlab/

[ROSE]

sorry for the inconvenience _ _

Answer 5

[ROSE]

can you explain the local maximum curvature method for finger vein extraction. i get the code for this extraction from miura_match method. but i can't understand it. please help me...

[Image Analyst]

It assumes that the vein will be a bright ridge or dark valley in an smooth background. Such a structure would have a high curvature in the intensity.

Answer 6

[lekshmi radhakrishnan]

i can't understand the repeated_line_tracking method. what you mean by the locus points? how it defines?