Intravenous grain extraction method based on maximal intra-neighbor difference vector diagram

Abstract

The invention discloses an intravenous grain extraction method based on maximal intra-neighbor difference vector diagram, which firstly calculates the maximal intra-neighbor difference and directions of all pixel points in an original intravenous image to construct a maximal intra-neighbor difference vector diagram, and further calculates the maximal intra-neighbor vector difference of the maximal intra-neighbor difference vector diagram, finally obtains proper thresholds from the histogram of the maximal intra-neighbor vector difference diagram and extracts the intravenous grain of the original image from the maximal intra-neighbor vector difference diagram. The invention avoids the image pretreatment process, also can eliminate the influence of intravenous image coherent features on grain extraction such as unevenness and fuzzy boundary, can completely extract the intravenous grain and ensures the grain to be even and have less noise points, and also greatly facilitates the following feature extraction.

Description

technical field [0001] The invention relates to a vein pattern extraction method, in particular to a vein pattern extraction method based on a maximum neighborhood interpolation vector diagram. Background technique [0002] Human body feature recognition is a technology that uses human physiological characteristics and behavior patterns for identification and authentication. Among them, vein recognition technology, as an emerging biometric technology, is non-contact, highly anti-interference and anti-counterfeiting. In recent years, more and more researchers have paid more and more attention to characteristics such as sex and liveness detection ability, and become a research hotspot in human body feature recognition. The technical basis of vein mode acquisition is near-infrared spectroscopy (NIRS) and near-infrared imaging. When 720nm-950nm near-infrared light irradiates a certain part of the human body, it will penetrate the surface of the skin, and when it enters the blood...

AI Technical Summary

Problems solved by technology

The other is to propose some new vein pattern extraction methods based on the characteristics of the vein pattern, such as using the repeated line tracing method to extract the vein pattern. This method randomly starts from different positions and detects local black spots. After the line is tracked coordinate by coordinate along the black line, the detection is repeated many times and all tracked black line trajectories are superimposed to realize the segmentation of the vein pattern, so this method can effectively extract the vein pattern from the less clear image, but for Among them, the thinner or shorter vein lines, because of the insufficient number of repeated detections on them, make this part of the vein lines cannot be effectively extracted and are easily lost; another example is a segmentation method based on local maximum curvature. In four different directions, the center points corresponding to the local maximum curvature values ​​of the cross-section are weighted and superimposed, and the remaining center points are connected to obtain the final vein pattern

The above algorithms have their own characteristics, but the vein patterns extracted by most of the algorithms are relatively thick, and it is difficult to segment thinner vein patterns, which have a certain distance from the real vein patterns, which is not conducive to the extraction of vein features. Therefore, will have a certain impact on the subsequent processing

Some scholars have proposed a skeleton extraction method based on the watershed algorithm for vein images. This algorithm firstly applies the watershed algorithm directly to the vein grayscale image to obtain vein lines containing many redundant ridges and noise, and then suppresses the noise through morphological processing. And delete redundant lines, but there are still some problems in this method: the vein lines containing floating endpoints cannot be extracted, and two vein lines that are closer to each other will be mistaken for one vein line

Some scholars have proposed a vein image segmentation algorithm based on the maximum intra-neighbor difference (MIND), the core of which is to make full use of the neighborhood information of the vein image and the newly designed distance function to calculate the MIND of the original image. image, and weighted and added to the original image after histogram correction to obtain the enhanced image, after that, the final segmentation result is obtained by calculating the mean image of the enhanced image and performing weighted comparison with the enhanced image. In addition, this method can be based on MIND The histogram of the image adaptively adjusts the segmentation parameters in the algorithm to improve the segmentation effect, but there are still some noises in the vein pattern obtained by this method

In response to these problems, some scholars proposed a vein image segmentation algorithm based on the distribution ratio of the direction field. This algorithm makes full use of the original vein image features and the spatial attributes of the direction field image, and uses the distribution ratio of the direction field as the segmentation method to distinguish veins from the background. According to the principle, better results can be achieved for textures with uneven thickness, but there are still more noises between similar textures, and a small amount of false textures may be caused due to the relatively concentrated noise

Images