Search

Image matrices are often transformed into vectors prior to feature extraction, which results in the curse of dimensionality when the dimensions of matrices are huge. In order to effectively deal with this problem, a new technique for two-dimensional(2D) Fisher discriminant analysis is developed in this paper. In the proposed algorithm, the Fisher criterion function is directly constructed in terms of image matrices. Then we utilize the Fisher criterion and statistical correlation between features to construct an objective function. We theoretically analyze that the proposed algorithm is equivalent to uncorrelated two-dimensional discriminant analysis in some condition. To verify the effectiveness of the proposed algorithm, experiments on ORL face database are made. Experimental results show that the performance of the proposed algorithm is superior to those of some previous methods in feature extraction. Moreover, extraction of image features using the proposed algorithm needs less time than that of classical linear discriminant analysis. [ABSTRACT FROM AUTHOR]

In this paper, an efficient iris segmentation method for recognition is described. The method is based on crossed chord theorem and zigzag collarette area. We select the zigzag collarette region as personal identification pattern, which can remove unnecessary areas and get good recognition rate. Zigzag collarette area is one of the most important parts of iris complex pattern. It is insensitive to the pupil dilation and not affected by the eyelid or eyelash since it is closed with the pupil. In our algorithm, we could avoid procedure for eyelid detection and searching the radius and the center position of the outer boundary between the iris and the sclera, which is difficult to locate when there is little contrast between iris and sclera regions. The method was implemented and tested using two iris database sets, i.e CASIA and SJTU-IDB, with different contrast quality. The experimental results show that the performance of the proposed method is encouraging and comparable to the traditional method. [ABSTRACT FROM AUTHOR]

Texture classification plays an important role in identifying objects. The fractal properties based on moment feature images for texture classification are investigated in this paper. The two-order moments of the image in small windows are used as feature images whose fractal dimensions are then computed and employed to classify the textures using support vector machines (SVMs). Experiments on several Brodatz nature images and four in-vivo B-mode ultrasound liver images demonstrate the effectiveness of the proposed algorithm. [ABSTRACT FROM AUTHOR]

Segmentation of connected handwritten Chinese characters is a very difficult task in document image analysis. In this paper, a novel algorithm based on stroke analysis and background thinning is proposed to segment connected handwritten Chinese characters. The feature points, viz. end points, fork points and corner points are detected in the thinned image. The segments between feature points are considered as substrokes and are extracted. Lengths of substrokes and the topological relations between them are employed to locate connected point. A new method based on background thinning is developed to decide a proper segmentation path. The experimental results show that satisfactory performance is achieved by the presented method for segmentation of connected handwritten Chinese characters. [ABSTRACT FROM AUTHOR]

An association rule typically strives for discovering a dependency among attributes with respect to the externally defined parameters like support threshold and confidence threshold. As an important database discovery method, the kernel of association rule mining is the acquisition of large itemsets. It is an important field of data mining to represent the support and confidence of items that are purchased together in supermarket domain. In this paper, a novel limited concept lattice is first proposed for the transaction data itemsets modeling. Concept lattice is a form of a concept hierarchy in which each node represents a subset of objects (extent) with their common properties (intent). The Hasse diagram of the lattice represents a generalization / specialization relationship between the concepts. Therefore, the lattice and Hasse diagram corresponding to a set of objects described by some properties can be used as an effective tool for symbolic data analysis and knowledge acquisition. Based on this lattice structure, an algorithm, LCLL, is presented to incrementally generate large itemsets visually. The algorithm works by means of attaching frequency information to each lattice node, the corresponding support measure can be obtained with the limited lattice. Besides, the edges in the Hasse diagram of the new lattice must be modified: the generator of a new node is always its child, and original parent of the generator is updated. When a node is deleted till the frequency value turns to zero, the node and the edges between its parents and children are not deleted, but tagged. The key point lies in adding edges when searching for the new node's parents, the large itemsets can be obtained by judging whether the cardinal and frequency value of the node exceeds the threshold or not. And accordingly, association rules can be identified. The approach is especially efficient when the database is dynamically updated (insertion, deletion or simultaneous insertion and deletion). Compared with K. Hu's approach [5], our algorithm generates all the association rules with much less time complexity. The time complexity of the proposed algorithm has a relationship of inverse proportion with the cardinal of the transactions, which means the applicability of the approach to the supermarket. [ABSTRACT FROM AUTHOR]