Signature recognition

About: Signature recognition is a research topic. Over the lifetime, 2138 publications have been published within this topic receiving 37605 citations.

A statistical approach for offline signature verification using local gradient features

Abstract: Signature is widely used as a means of personal verification which emphasizes the need for a signature verification system. Often the single signature feature may produce unacceptable error rates. Features used in this method are mainly local key-point feature that deals with the orientation around each key-point. Before extracting the features, preprocessing of a scanned image is necessary to isolate the region of interest part of a signature and to remove any spurious noise present. The system is initially trained using a database of signatures obtained from those individuals whose signatures are to be authenticated by the system. For extracting the feature, key-points of the image are detected. For each point, orientation around the point is calculated as the feature. By matching the features of sample signature and testing signature decision is taken. If a query signature is in the acceptance range then it is an authentic signature, otherwise it is a forged signature.

Action recognition by graph embedding and temporal classifiers

Abstract: With the improved accessibility to an exploding amount of video data and growing demand in a wide range of video analysis applications, videobased action recognition becomes an increasingly important task in computer vision. Unlike most approaches in the literature which rely on bagof-feature methods that typically ignore the structural information in the data, in this monograph we incorporate the spatial relationship and the time stamps in the data in the recognition and classification processes. We capture the spatial relationships in the subject performing the action by representing the actor’s shape in each frame with a graph. This graph is then transformed into a vector of real numbers by means of prototypebased graph embedding. Finally, the temporal structure between these vectors is captured by means of sequential classifiers. The experimental results on a well-known action dataset (KTH) show that, although the proposed method does not achieve accuracy comparable to that of the best existing approaches, these embedded graphs are capable of describing the deformable human shape and its evolution over time. We later propose an extended hidden Markov model, called the hidden Markov model for multiple, irregular observations (HMM-MIO), capable of fusing spatial information provided by graph embedding and the textural information of STIP descriptors. Experimental results show that recognition accuracy can be significantly improved by combining the spatiotemporal features with the structural information obtaining higher accuracy than from either separately. Furthermore, HMM-MIO is applied to the task of joint action segmentation and classification over a concatenated version of the KTH action dataset and the challenging CMU multi-modal activity dataset. The achieved accuracies proved comparable to or higher than state-of-the-art approaches and show the usefulness of the proposed model also for this task. The next and most remarkable contribution of this dissertation is the creation of a novel framework for selecting a set of prototypes from a labelled graph set taking class discrimination into account. Experimental results show that such a discriminative prototype selection framework can achieve superior results, not only for the task of human action recognition, but also in the classification of various structured data such as letters, digits, drawings, fingerprints compared to other well-established prototype selection approaches. Lastly, we change our focus from the forementioned problems to the recognition of complex event, which is a recent area of computer vision expanding the traditional boundaries of visual recognition. For this task, we have employed the notion of concept as an alternative intermediate representation with the aim of improving event recognition. We model an event by a hidden conditional random field and we learn its parameters by a latent structural SVM approach. Experimental results over video clips from the challenging TRECVID MED 2011 and MED 2012 datasets show that the proposed approach achieves a significant improvement in average precision at a parity of features and concepts.

Multimodal biometric system using index based algorithm for fast search

Abstract: Establishing the identity of a person with the use of individual biometric features has become the need for the present technologically advancing world. Due to rise in data thefts and identity hijacking, there is a critical need for providing user security using biometric authentication techniques. A unimodal biometric system is known to have many disadvantages with regard to accuracy, reliability and security. Multimodal biometric systems combine more than one biometric trait to identify a person for enhanced security. The proposed Multimodal biometric system combines three biometric traits for individual authentication namely Face, Fingerprint and Voice and implements indexing algorithm for faster searching and recognition of a person. The fuzzy nature of biometric data and the presence of varied degree of dimensionality hinder present search algorithms depending on sorting. Index based algorithm is used to reduce the search time and also to improve the speed and performance of multimodal biometric system. MapReduce is used for analyzing and processing big data sets that cannot fit into memory.

A Robust On-line Signature Verification System

Abstract: This paper proposes a robust on-line signature verification system based on a new segmentation method and fusion scheme. The proposed segmentation method resolves the problem of segment-to-segment comparison where the variation between reference signature and input signature causes the errors in the location and the number of segments. In addition, the fusion scheme is adopted, which discriminates genuineness by calculating each feature vector`s fuzzy membership degree yielded from the proposed segmentation method. Experimental results show that the proposed signature verification system has lower False Reject Rate(FRR) for genuine signature and False Accept Rate(FAR) for forgery signature.