Automatic Signature detection using image processing

Published on 03 Jul 2021. Written by Shirin Tikoo

The way that the signature is broadly utilized as a method for individual distinguishing proof apparatus for people necessitates. Confirmation of signature can be performed either Offline or Online dependent on the application. Anyway, human signatures can be dealt with an image and perceived utilizing computer vision and neural network methods. Signature verification and recognition is an innovation that can improve security in our everyday exchanges held in the public arena. This system is reasonable for different applications, for example, bank exchanges, travel papers with great confirmation results, and so on.

Read more..

SLNOTE


The structure of any signature verification system generally requires a few important steps:

Image Scanning and Reading- Image of the signature is scanned, filtered and put away as JPEG(Joint photographic expert group) during examining, scanning and storing procedures to prevent image loss data.

Image Pre-processing- Image preprocessing includes image smoothening, noise removal, border elimination, and image normalization.

Image Smoothening- During the picture filtering, we can get some commotion (noise) over the image. So to expel the commotion, image smoothing is being performed.

Grayscale and Binary Scale conversion- Grayscale expresses the hues i.e colors going from dark to white though Binary scale expresses commonly two value i.e high contrast and for appropriate division, we utilize the binary scale.

Border Elimination or Edge Detection- The Border Removal image upgrade consequently expels the border around the image. For instance, a black border around the edges of scanned documents. Evacuating an image's border will influence the size and dimensions of the image,  it won't scale or change the resolution of the image.

SLLATEST
Image Normalization: In image processing, normalization is a procedure that changes the scope of pixel intensity values. Applications incorporate photos with a poor difference because of glare. Normalization is now and then called contrast stretching or histogram stretching.

Feature extraction: The feature extraction process incorporates the extraction of the input image highlights. Feature extraction is an essential step for signature recognition, which consequences for the structure and execution of the classifier successfully.

Comparison process & performance evaluation

This stage involves the comparison of the processed image and the actual image to get the results.

MATLAB environment- MATLAB tool is used for image processing and for analyzing the final results.

Database- We have to create a little database for putting the database.

Project Implementation-

There are a couple of steps for accomplishing the conclusive outcome and the means are:

1. The initial step is to capture and scan the input image and start the image pre-processing steps.
2. The subsequent stage is to convert the input image into a grayscale image.
3. After the conversion of the input image, we will perform edge detection or border elimination process.
4. Presently we will do the comparison process & performance evaluation of the input image and reference image from the database.
5. Finally, we get the outcome that if the signature is fake or the genuine one.

Software requirement-

Programming language - MATLAB and any database.

SLDYK
Kit required to develop Automatic Signature detection using image processing :

• No kit

Technologies you will learn by working on Automatic Signature detection using image processing :

• Image Processing
• Computer Vision
• MATLAB

Any Questions?