Hi Utkarsh,
Kindly Update the numpy library
It messing up with my environment variable

Hi,
I am building the fingerprint verification app but your enhance algorithm takes 2.5s. How can I to reduce processing time?
Thank you.

Hey Utkarsh, it would be helpful if we will be able to enhance fingerprints taken by camera. Attaching a sample image

Could you please tell me how can I revert the images from the current result to white background and black ridges fingerprint?
Thank you very much!

Hi Utkarsh,

You did a great job. The fingerprint image after enhancement look great! I was trying to convert your python code into C#. I am using emgu.CV, the closest method to achieve the functionality of ndimage.convolve was using cv2.filter2D because this method is supported in C#.

The python code is as follow :
Gx = ndimage.convolve(Gxx,f)

and C# code is as follow

 ConvolutionKernelF GxxKernal = new ConvolutionKernelF(f);
        f= f.Flip(FlipType.Both);
        CvInvoke.Filter2D(ToMatrix(Gxx), Gx, f, new Point(-1, -1), 0, BorderType.Reflect);

by using this parameter for Filter2D in openCV, I can get a close result to python however there are not equal.
Maybe Filter2D in OpenCV is not capable of producing the same result as in ndimage.convolve, do you know is there any other method in opencv that can produce the same convolution result as in ndimage.convolve? Thanks in advance!

I tried to input an image, but the image is kind of buggy. In the original image, there are some white lines in the centre of the fingerprint. This feature feels as if there is some glitch when it comes to enhancing the picture. The input image is in the bmp format.
This is the input image:

And this is the output image:

I am getting this error -

AttributeError: partially initialized module 'fingerprint_enhancer' has no attribute 'enhance_Fingerprint' (most likely due to a circular import)

Hi @Utkarsh-Deshmukh,

I was trying to run your code on fingerprint patches resized to 256x256. The ridges maps aren't formed completely(might be due to pores). There is also a black border along the edges but I need the ridge map for the complete image.

I have attached the resulting ridge map below along with the original image. Could you please guide me on resolving these issues?

Original Image (96x96)

Enhanced Ridge Map Image (256x256)

Thank you

Hi,

I have converted the code to pure c# however there are some discrepancy on the final image. So I try to visualize the orientation image for Sin, Cos and etc

I found out the Cos image generated from c# is as follow

Generated by Python code

Orientation (c#)

Orientation (Python)

Here is the original fingerprint

Enhance result from c#

Enhance result from Python

Could it be the different in the Cos orientation image as generated by both C# and Python that cause the subsequence failure in the final enhance image?

in ridge_orient.py :

rows,cols = im.shape;
sze = np.fix(6*gradientsigma); # this line
if np.remainder(sze,2) == 0:
sze = sze+1;

im a novice at python, installed the enhancer and after getting the script to show the enhance image its just a a grey screen wait for over an hour and retried multiple times not sure whats going wrong

Hello Utkarsh, maybe the problem is so silly and the solution is so simple but i'cant execute main_enhancement.py in anaconda. The error is: AttributeError: module 'scipy.ndimage' has no attribute 'imread'. I try to change the ndimage to misc and imageio but it doesn't help, error is the same. While i've been search for the solution (especially on stackoverflow) i saw that someone else's problem solved with importing pillow, so i tried it too but nothing changed. Is this problem occurs because of the version differences? What do you think?

Hii Utkarsh,
I am running it for my palm images and for some of the images it works perfectly well but for some of the images there is an error.
Error is: index 0 is out of bounds for axis 0 with size 0
in function: __ridge_filter
In line : sigmax = 1 / unfreq[0] * self.kx
When i try to debug it i found that variable unfreq has shape as (0,).
I don't know why for some of the images it is working but for other it is not.
I am also attaching image for your reference.
One more thing i have resized my current image and after resizing it is not working and if I pass the image without resizing it is working fine. But again i did'nt get the reason can you explain and tell me how to solve it.
The following is without resizing

The following is after resizing.

Thanks
Kapil singla

I tried converting some more images, and some images still behave buggy. I know it is very hard to solve, but are there any approaches to solve it? Also, I found out the later version of the code still somehow can't work as intended and even makes it worse.
Original:

pip:

Raw code:

Hi!
I was just trying to run this code. I got an error that there's no attribute "int" for module numPy. I converted that to normal int() function. Now I'm getting a new error "TypeError: only size-1 arrays can be converted to Python scalars" on the line "cropsze = int(np.fix(rows // np.sqrt(2)))". Can you tell me what I'm doing wrong? Should i change my python version or opencv version?

Hi, I am using this library on my fingerprint dataset and i get this error:

`in ridge_filter(im, orient, freq, kx, ky)
75 # orientations in 'angleInc' increments.
76
---> 77 sigmax = 1/unfreq[0]*kx;
78 sigmay = 1/unfreq[0]*ky;
79

IndexError: index 0 is out of bounds for axis 0 with size 0`

Can you help me?

Hello author, I had a problem why I ran according to your method 1, a problem appeared - this dll could not be found.

Traceback (most recent call last):
File "example.py", line 9, in
from src.FingerprintImageEnhancer import FingerprintImageEnhancer
File "C:\Users\wrx\Desktop\SR\image-enhance\Fingerprint-Enhancement-Python-develop\src\FingerprintImageEnhancer.py", line 10, in
from scipy import signal
File "D:\Wangluohuanjing\Anaconda3\lib\site-packages\scipy\signal_init_.py", line 322, in
from .filter_design import *
File "D:\Wangluohuanjing\Anaconda3\lib\site-packages\scipy\signal\filter_design.py", line 18, in
from scipy import special, optimize, fftpack
File "D:\Wangluohuanjing\Anaconda3\lib\site-packages\scipy\optimize_init_.py", line 387, in
from .minimize import *
File "D:\Wangluohuanjing\Anaconda3\lib\site-packages\scipy\optimize_minimize.py", line 30, in
from .trustregion_constr import minimize_trustregion_constr
File "D:\Wangluohuanjing\Anaconda3\lib\site-packages\scipy\optimize_trustregion_constr_init.py", line 4, in
from .minimize_trustregion_constr import minimize_trustregion_constr
File "D:\Wangluohuanjing\Anaconda3\lib\site-packages\scipy\optimize_trustregion_constr\minimize_trustregion_constr.py", line 4, in
from scipy.sparse.linalg import LinearOperator
File "D:\Wangluohuanjing\Anaconda3\lib\site-packages\scipy\sparse\linalg_init.py", line 117, in
from .eigen import *
File "D:\Wangluohuanjing\Anaconda3\lib\site-packages\scipy\sparse\linalg\eigen_init.py", line 11, in
from .arpack import *
File "D:\Wangluohuanjing\Anaconda3\lib\site-packages\scipy\sparse\linalg\eigen\arpack_init.py", line 22, in
from .arpack import *
File "D:\Wangluohuanjing\Anaconda3\lib\site-packages\scipy\sparse\linalg\eigen\arpack\arpack.py", line 45, in
from . import _arpack
ImportError: DLL load failed: 找不到指定的程序。

Hi, Utkarsh
When I saw the code "orientim = np.pi/2 + np.arctan2(sin2theta,cos2theta)/2;" in ridge_orient.py,
I was confused about why need to add pi/2. Then, I drew a directional field on a img,but it looked irregular.
Looking forward to your reply.

I want to run this project but failed many times

Hi Utkarsh,
when i try and run the script no matter the dpi of the original image, the new images DPI is 96 and it's resolution is too low for the program we use to scan fingerprints, is there any way to fix this?

Cheers,
Bradley

Hi Utkarsh,
When I run your code,
I get the following error

Macintosh:src saikiranoruganti$ python main_enhancement.py

gabor_filter = np.array(np.zeros((180/angleInc,filt_rows,filt_cols)));

TypeError: 'float' object cannot be interpreted as an integer
Macintosh:src saikiranoruganti$ 180/angleInc,filt_rows,filt_cols

It seems the floats are unable to pass through np.zeros. Any directions on how to fix this?

Here is the result generated from
C#

Python

The ridges generated from original image are more thicker and connected while the one using openCV replacement function are thinner and disconnected.

There are a lot of useful library found in python but didn't provided by C# so I have to make some changes on certain function call, mainly using opencv to replace sciPy's function.

In the ridge_orient() function
I replace

        gauss = cv2.getGaussianKernel(np.int(sze),self.gradient_sigma)
        f = gauss * gauss.T

        fy,fx = np.gradient(f)    

        Gx = signal.convolve2d(self._normim, fx, mode='same')
        Gy = signal.convolve2d(self._normim, fy, mode='same')

        Gxx = ndimage.convolve(Gxx,f)
        Gyy = ndimage.convolve(Gyy,f)
        Gxy = 2*ndimage.convolve(Gxy,f)

with

        gauss = cv2.getGaussianKernel(np.int(sze),self.gradient_sigma)
        f = gauss * gauss.T

        fx=cv2.Sobel(f,cv2.CV_64F,1,0,3)
        fy=cv2.Sobel(f,cv2.CV_64F,0,1,3)

       Gx = cv2.filter2D(self._normim,-1,cv2.flip(fx,-1),borderType=cv2.BORDER_REFLECT)
        Gy = cv2.filter2D(self._normim,-1,cv2.flip(fy,-1),borderType=cv2.BORDER_REFLECT)

        Gxx = cv2.filter2D(Gxx,-1,cv2.flip(f,-1),borderType=cv2.BORDER_REFLECT)
        Gyy = cv2.filter2D(Gyy,-1,cv2.flip(f,-1),borderType=cv2.BORDER_REFLECT)
        Gxy = 2*cv2.filter2D(Gxy,-1,cv2.flip(f,-1),borderType=cv2.BORDER_REFLECT)

__frequest()

I replace

rotim = scipy.ndimage.rotate(blkim, orient / np.pi * 180 + 90, axes=(1, 0), reshape=False, order=3,
                                    mode='nearest')

with

       ROT_mat = cv2.getRotationMatrix2D((cols/2,rows/2),orient/np.pi*180 + 90,1)
        rotim = cv2.warpAffine(blkim,ROT_mat,(cols,rows))

What is the reason that cause such discrepancy? Any Idea to make the ridge thicker and connected? Thanks in advance

Hi there!
I'm trying to run this project, but there are some errors. I have tried to fix them but anyway can't run. I attach the 1st screenshot of the first error. This error I've fixed placing '//' instead of '/'. Then was similar error and after was error like on 2nd screenshot.
I'm on windows 10 and using Anaconda 2 spyder (Python 3.7)