[17] Let �� > 0 be a given constant, then the filtering error system Gefitinib clinical trial (10) is said to have a finite-frequency l2 gain ��, if inequality��k=0��e(k)Te(k)�ܦ�2��k=0��w(k)Tw(k)(12)holds for all solutions of Equation (10) with w(k) l2 such that the following holdFor the low-frequency range |��| < l��k=0��(��(k+1)?��(k))(��(k+1)?��(k))T��(2sin?l2)2��k=1�ަ�(k)��(k)T(13)For the middle-frequency range 1 �� �� �� 2ej?w��k=0��(��(k+1)?ej?1��(k))(��(k+1)?e?j?2��(k))T��0(14)where w = (2 ? 1)/2.For the high-frequency range |��| �� h��k=0��(��(k+1
In recent years, palmprint recognition has drawn widespread attention from researchers. Generally, palmprint recognition involves using the person’s palm to identify who the person is or verify whether the person is ��whom he claims to be��.
Some previous researches have shown that, compared with fingerprints or iris- based personal biometrics systems, palmprint-based biometric systems have several special advantages such as rich features, less distortion and easy self-positioning [1�C6]. And, it can also obtain high accurate recognition rate with fast processing speed [2�C6]. For the aforementioned reasons, nowadays research on palmprint recognition is becoming more and more active [5,6].Roughly speaking, the techniques of palmprint recognition can be divided into two categories, i.e., 2-D based [5] and 3-D based [7], respectively. As their name suggests, 2-D based palmprint recognition techniques capture a 2-D image of the palm surface and use it for feature extraction and matching, while 3-D based techniques capture the 3-D depth information for recognition.
As noted in the literature [7], 3-D palmprint recognition techniques offer some special advantages. For example, they are robust to illumination variations, contaminations and spoof attacks. However, the cost of 3-D data acquisition devices is high, which limits the usage of 3-D palmprint recognition techniques [7]. Therefore, 2-D palmprint recognition has drawn more attention in the past decade [6]. In this paper, we also focus on it.It is well known that the palm contains rich features such as minutiae, ridges, principal lines and creases. In a high-resolution (500 ppi or higher) palmprint image, all features mentioned above can be extracted. Recently, there have been several works related to high-resolution palmprint recognition [8,9].
In fact, most high-resolution palmprint recognition techniques are mainly developed for forensic applications as about 30 percent of the latents recovered from crime scenes are from palms [9]. On the Batimastat other hand, for civil applications, the technique of low-resolution (about 100 ppi) palmprint recognition customer reviews is enough for robust personal authentication. In this paper, our work also belongs to the low-resolution palmprint recognition category. In a low-resolution palmprint image, only principal lines and creases can be extracted to construct features.