Kuanquan Wang

Fisherpalms based palmprint recognition

In this paper, a novel method for palmprint recognition, called Fisherpalms, is proposed. In this method, each pixel of a palmprint image is considered as a coordinate in a high-dimensional image space. A linear projection based on Fishers linear discriminant is used to project palmprints from this high-dimensional original palmprint space to a significantly lower dimensional feature space (Fisherpalm space), in which the palmprints from the different palms can be discriminated much more efficiently. The relationship between the recognition accuracy and the resolution of the palmprint image is also investigated. The experimental results show that, in the proposed method, the palmprint images with resolution 32 × 32 are optimal for medium security biometric systems while those with resolution 64 × 64 are optimal for high security biometric systems. High accuracies (>99%) have been obtained by the proposed method and the speed of this method (responding time ≤ 0.4 s) is rapid enough for real-time palmprint recognition.

Computerized tongue diagnosis based on Bayesian networks

Tongue diagnosis is an important diagnostic method in traditional Chinese medicine (TCM). However, due to its qualitative, subjective and experience-based nature, traditional tongue diagnosis has a very limited application in clinical medicine. Moreover, traditional tongue diagnosis is always concerned with the identification of syndromes rather than with the connection between tongue abnormal appearances and diseases. This is not well understood in Western medicine, thus greatly obstruct its wider use in the world. In this paper, we present a novel computerized tongue inspection method aiming to address these problems. First, two kinds of quantitative features, chromatic and textural measures, are extracted from tongue images by using popular digital image processing techniques. Then, Bayesian networks are employed to model the relationship between these quantitative features and diseases. The effectiveness of the method is tested on a group of 455 patients affected by 13 common diseases as well as other 70 healthy volunteers, and the diagnostic results predicted by the previously trained Bayesian network classifiers are reported.

Wavelet-based cascaded adaptive filter for removing baseline drift in pulse waveforms

This paper presents an energy ratio-based method and a wavelet-based cascaded adaptive filter (CAF) for detecting and removing baseline drift from pulse waveforms. Experiments on 50 simulated and five hundred real pulse signals demonstrate that this CAF outperforms traditional filters both in removing baseline drift and in preserving the diagnostic information of pulse waveforms.

Bidirectional PCA with assembled matrix distance metric for image recognition

Principal component analysis (PCA) has been very successful in image recognition. Recent research on PCA-based methods has mainly concentrated on two issues, namely: 1) feature extraction and 2) classification. This paper proposes to deal with these two issues simultaneously by using bidirectional PCA (BD-PCA) supplemented with an assembled matrix distance (AMD) metric. For feature extraction, BD-PCA is proposed, which can be used for image feature extraction by reducing the dimensionality in both column and row directions. For classification, an AMD metric is presented to calculate the distance between two feature matrices and then the nearest neighbor and nearest feature line classifiers are used for image recognition. The results of the experiments show the efficiency of BD-PCA with AMD metric in image recognition

The bi-elliptical deformable contour and its application to automated tongue segmentation in Chinese medicine

Automated tongue image segmentation, in Chinese medicine, is difficult due to two special factors: 1) there are many pathological details on the surface of the tongue, which have a large influence on edge extraction; 2) the shapes of the tongue bodies captured from various persons (with different diseases) are quite different, so they are impossible to describe properly using a predefined deformable template. To address these problems, in this paper, we propose an original technique that is based on a combination of a bi-elliptical deformable template (BEDT) and an active contour model, namely the bi-elliptical deformable contour (BEDC). The BEDT captures gross shape features by using the steepest decent method on its energy function in the parameter space. The BEDC is derived from the BEDT by substituting template forces for classical internal forces, and can deform to fit local details. Our algorithm features fully automatic interpretation of tongue images and a consistent combination of global and local controls via the template force. We apply the BEDC to a large set of clinical tongue images and present experimental results.

Fast neighborhood component analysis

Distance metric is of considerable importance in varieties of machine learning and pattern recognition applications. Neighborhood component analysis (NCA), one of the most successful metric learning algorithms, suffers from the high computational cost, which makes it only suitable for small-scale classification tasks. To overcome this disadvantage, we proposed a fast neighborhood component analysis (FNCA) method. For a given sample, FNCA adopts a local probability distribution model constructed based on its K nearest neighbors from the same class and from the different classes. We further extended FNCA to nonlinear metric learning scenarios using the kernel trick. Experimental results show that, compared with NCA, FNCA not only significantly increases the training speed but also obtains higher classification accuracy. Furthermore, comparative studies with the state-of-the-art approaches on various real-world datasets also verify the effectiveness of the proposed linear and nonlinear FNCA methods.

Orientation selection using modified FCM for competitive code-based palmprint recognition

Coding-based methods are among the most promising palmprint recognition methods because of their small feature size, fast matching speed and high verification accuracy. The competitive coding scheme, one representative coding-based method, first convolves the palmprint image with a bank of Gabor filters with different orientations and then encodes the dominant orientation into its bitwise representation. Despite the effectiveness of competitive coding, few investigations have been given to study the influence of the number of Gabor filters and the orientation of each Gabor filter. In this paper, based on the statistical orientation distribution and the orientation separation characteristics, we propose a modified fuzzy C-means cluster algorithm to determine the orientation of each Gabor filter. Since the statistical orientation distribution is based on a set of real palmprint images, the proposed method is more suitable for palmprint recognition. Experimental results indicate that the proposed method achieves higher verification accuracy while compared with that of the original competitive coding scheme and several state-of-the-art methods, such as ordinal measure and RLOC. Considering both the computational complexity and the verification accuracy, competitive code with six orientations would be the optimal choice for palmprint recognition.

Coarse iris classification using box-counting to estimate fractal dimensions

This paper proposes a novel algorithm for the automatic coarse classification of iris images using a box-counting method to estimate the fractal dimensions of the iris. First, the iris image is segmented into sixteen blocks, eight belonging to an upper group and eight to a lower group. We then calculate the fractal dimension value of these image blocks and take the mean value of the fractal dimension as the upper and the lower group fractal dimensions. Finally, all the iris images are classified into four categories in accordance with the upper and the lower group fractal dimensions. This classification method has been tested and evaluated on 872 iris cases, and the proportions of these categories in our database are 5.50%, 38.54%, 21.79%, and 34.17%. The iris images are classified with two algorithms, the double threshold algorithm, which classifies iris images with an accuracy of 94.61%, and the backpropagation algorithm, which is 93.23% accurate. When we allow for the border effect, the double threshold algorithm is 98.28% accurate.

Tongue image analysis for appendicitis diagnosis

Medical diagnosis using the tongue is a unique and important diagnostic method of traditional Chinese medicine (TCM). However, the clinical applications of tongue diagnosis have been limited due to two factors: (1) tongue diagnosis is usually based on the capacity of the eye for detailed discrimination; (2) the correctness of tongue diagnosis depends on the experience of physicians; and (3) traditional tongue diagnosis is always dedicated to the identification of syndromes other than diseases. To address these problems, in this paper, we present a tongue-computing model (TCoM) for the diagnosis of appendicitis based on quantitative measurements that include chromatic and textural metrics. These metrics are computed from true color tongue images by using appropriate techniques of image processing. Applying our approach to clinical tongue images, the experimental results are encouraging.

The relative distance of key point based iris recognition

Iris recognition has received increasing attention in recent years as a reliable approach to human identification. This paper makes an attempt to analyze the local feature structure of iris texture information based on the relative distance of key points. When preprocessed, the annular iris is normalized into a rectangular block. Multi-channel 2-D Gabor filters are used to capture the iris texture. In every filtered sub-image, we extract the points that can represent the local texture most effectively in each channel. The barycenter of these points in each channel is called the key point and a group of key points are obtained. Then, the distance between the center of key points of each sub-image and every key point is called relative distance, which is regarded as the iris feature vector. Iris feature matching is based on the Euclidean distance. Experimental results on public and private databases show that the performance of the proposed method is encouraging.