A Method of Edge Extraction and Direction Correction of Finger Vein Image

Abstract

The invention provides a method for edge extraction and direction correction of a finger vein image, comprising the following steps: performing cutting to remove some inherent equipment interference; using a sobel horizontal operator to detect the upper and lower contours of the finger to obtain the horizontal and vertical coordinates of the upper and lower edge points; Get the edge of noise / interference and remove the noise; perform cubic curve fitting on the upper edge point and lower edge point respectively, and interpolate the missing part of the edge to obtain the upper edge curve C up and lower edge curve C down ; via C up and C down Solve for the axial centerline L of the finger mid ;according to the image centerline and L mid The intersection point determines the center of rotation, according to the image centerline and L mid The angle of rotation determines the rotation angle, and the image processing result is obtained after the original image is rotated and corrected according to the rotation center and angle. The invention self-adaptively detects edges and performs direction correction, can adapt to various images with unclear, blurred edges, low contrast and noise, and provides high-quality images for further feature value extraction and feature value comparison.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to a method for edge extraction and direction correction of a finger vein image. Background technique [0002] The camera captures the finger vein image through the glass window of the filter, and the size of the window determines the amount of information in the finger vein image. Existing finger vein acquisition and processing devices often only care about part of the information in the finger veins, and often acquire finger vein images through a small window. The vein information in the finger vein picture is related to the position and angle of the finger. The thickness of the fingers of different people is different. The finger vein information contained in the image has a certain randomness, and the blood vessels with the most information are the first, The two joints can't be determined either. Expanding the window can obtain more comprehensive finger vein informat...

AI Technical Summary

Problems solved by technology

Expanding the window can obtain more comprehensive finger vein information in addition to blood vessel information, such as finger direction, finger contour, knuckle position, etc., which is more conducive to ensuring the accuracy of identity authentication, but the edge extraction and direction of finger vein image Correction puts forward higher requirements

Images