In order to enhance the fingerprint image such that ridges of the fingerprint can be connected, consider using image thresholding techniques for initially enhancing the contrast of the image and then using edge detection to detect ridges appropriately.
Here is an example which follows:
- 'adapthisteq' function to obtain an image with high contrast between ridges and non-ridges than initial image. This is based on histogram equalization technique of image thresholding.
- 'edge' function is used to detect the edges of the ridges from the high contrast image. Here I have used 'Canny' method of edge detection.
- 'strel' and 'imdilate' functions have been used to emphasize the ridges based on a particular size of structuring element and connecting broken ridges.
% Assuming 'img' to be the variable containing the image
% Histogram equalization
grayImg = rgb2gray(img);
enhancedImg = adapthisteq(grayImg, "Distribution","exponential", "Alpha",0.9);
% Edge detection and Binarization
edges = edge(enhancedImg, 'Canny');
binaryImg = imbinarize(enhancedImg);
se = strel('disk',0);
dilatedImg = imdilate(binaryImg, se);
% The parameter arguments of the provided code can be modified according to
% custom requirements.
I am able to achieve the following enhanced images after histogram equalization and edge detection and binarization:
The following MATLAB Answer also mentions some research papers for fingerprint recognition methods research papers: https://www.mathworks.com/matlabcentral/answers/113804-how-to-segment-a-fingerprint-image
The following MathWorks documentations can be referred to know more:
'adapthisteq': https://www.mathworks.com/help/images/ref/adapthisteq.html
'Image Thresholding': https://www.mathworks.com/discovery/image-thresholding.html
Thanks.
