Region of interest extraction for palmprint and palm vein recognition

The ROI (region of interest) extraction is the key step in palmprint or palm vein recognition, which is very important for the subsequent feature extraction and recognition. In this paper, the ROI extraction method for palmprint and palm vein recognition is mainly studied. Firstly, the preprocessing operation of palmprint and palm vein is carried out by using binary and morphological denoising technology, then the ROI regions are located and extracted based on the maximum inscribed circle and centroid methods. Finally, the algorithms are tested on PUT palm vein database and CASIA database, the experimental results show that the methods of this paper have a good effect, which are feasibility and validity. Furthermore, an online ROI extraction simulation system is designed in this article. The palmprint or palm vein image can be obtained in real-time by using the camera through the system, then the ROI can be extracted. The simulation system is intuitive and easy to operate, which provides a reliable experimental platform for the study of palmprint and palm vein recognition technology.

I. Introduction

Paimprint recognition [1]–[2] is based on the effective texture information on the palm of the hand for identification, and the palm vein recognition [3]–[4] is according to the information of vein blood vessel under the palm skin layer. The ROI extraction is a hot issue for the palmprint and palm vein identification. The ROI extraction refers to carrying out a series of adjustment and key points location for different palmprint and palm vein images, then the effective area of center is selected to extract features, and final matching is carried out for the recognition. This central region is usually called the region of interest (ROI), for the palmprint and palm vein image of the same palm, the location of ROT should be the same. The purpose of ROT location and selection is to do the feature area normalization of the different palmprint and palm vein, so the influence of adverse factors will be eliminated, and the sub image including rich information of palm print or palm vein is extracted, which is convenient for the subsequent feature extraction and matching. There are a lot of ROT extraction methods at present [6]–[9]. The ROT extraction methods of palmprint and palm vein based on the maximum inscribed circle [8] and centroid methods [4] are mainly studied in this paper. The maximum inscribed circle method is that the center and radius would be determined in the palm of the hand, and the largest inscribed circle of all palm area is obtained, then the angle rotation correction of the palm is carried out. After that the maximum inscribed square is cut out in the inscribed circle, which is finally normalized to the size of 128×128 ROT image. The method based on the centroid extraction is that the centroid of the palm image is obtained, and the rectangle sub image which size is 256×256 would be intercepted by using centroid as the center. The experimental results based on PUT vein database [10] and CASIA database [11] show that the method in this paper has good application prospects. In addition, an online simulation system which is easy to use and operate is designed in this paper to obtain the image of palmprint or palm vein and extract ROT in real-time, and it will provide a good research platform for algorithm verification.