Improving Iris Recognition Accuracy via Cascaded Classifiers

Abstract

Abstract. As a reliable approach to human identification, iris recognition has received increasing attention in recent years. In the literature of iris recognition, local feature of image details has been verified as an efficient iris signature. But measurements from minutiae are easily affected by noises, which greatly limits the system's accuracy. When the matching score between two intra-class iris images is near the local feature based classifier's (LFC) decision boundary, the poor quality iris images are usually involved in matching. Then a novel iris blob matching algorithm is resorted to make the recognition decision which is more robust than the LFC in the noisy environment. The extensive experimental results demonstrate that the cascading scheme significantly outperforms individual classifier in terms of accuracy and robustness.

A Review on Biometric Recognition

“…Chen et al, (2006) has proposed Daugman's 2-D Gabor filter with quality measure enhancement [22] and Du et al, (2006) proposed method of recognition using 1-D local texture patterns [23].…”

Iris Recognition based on Wavelets

“…The feature values are the mean and the average absolute deviation of the magnitude of each 8x8 block in the filtered image. The method by Li Ma was further improved by Zhenan Sun where in the local feature based classifier was combined with an iris blob matcher [13]. The blob matching aimed at finding the spatial correspondences between the blocks in the input image and that in the stored model.…”

Computationally Efficient Serial Combination of Rotation-Invariant and Rotation Compensating Iris Recognition Algorithms

“…Sun et al (Sun et al, 2005) "cascade" two feature types employing global features only in addition to a Daugman-like approach if the matching value of the latter is in a questionable range. Also Zhang et al (Zhang et al, 2004) use a similar strategy while interchanging the role of the global and the local features.…”