Topic
Signature recognition
About: Signature recognition is a research topic. Over the lifetime, 2138 publications have been published within this topic receiving 37605 citations.
Papers published on a yearly basis
Papers
More filters
••
TL;DR: The approach of human identity and gender recognition is presented using Model based features extraction and SURF for matching along with SVM and MDA algorithm.
Abstract: The identification through biometric is a better way because it associate with individual not with information passing from one place to another. There are numerous biometric measures which can be used to help derive an individual identity. It is the biometric process and has many advantages over other biometric traits such as face, iris, fingerprint, palm print, etc. Most current approaches make the unrealistic assumption that persons walk along a fixed direction or a pre-defined path. Gait is the manner or style of moving on foot. Human Gait recognition identifies the individuals by the way in which they walk. Recognition of an individual is an important task to identify people. A gait sequence is collected from arbitrary walking directions. In this paper we present the approach of human identity and gender recognition using Model based features extraction and SURF for matching along with SVM and MDA algorithm.
3 citations
••
02 Oct 2018TL;DR: This work proposes a new method to recognize handwritten signature in an offline manner using the centroid of two local binary vectors, the horizontal vector and the vertical vector.
Abstract: The Handwritten Signature is a special sign used by humans and may contain letters, curves, or both. The main usage of handwritten signature is a proof of identification, especially when dealing with official documents and treatments. To recognize a signature means to identify the person who uses this sign. Signature recognition has many applications, such as: transactions and checks in banking systems, forensic caseworks, personal authentication and verification. This work proposes a new method to recognize handwritten signature in an offline manner. The centroid of two local binary vectors, the horizontal vector and the vertical vector are calculated. Three different tests are accomplished for this method. Soft evaluation test, Hard evaluation test, and a Combined test. The gained results from the three of these tests are encouraging. It reached for what is so called the Combined test to 94.8275% of success rate for 928 digital images of handwritten signatures with processing time for single sample reaches to 0.146 in milliseconds.
3 citations
01 Jan 2008
TL;DR: A rotational as well as translational-invariant scheme by which the problem of real-time palmprint verification can be overcome while preprocessing the image before the feature extraction of the palmprint.
Abstract: Biometric identification is an emerging technology that can solve security problems in our networked society. As the important implementation of biometric technology, palmprint verification is one of the most reliable personal identification methods. Human palmprint recognition has become an active area of research over the last decade. In this paper, a new efficient approach to the palmprint preprocessing scheme is presented. In real-time palmprint verification the input subject to the scanner for image acquisition may suffer rotational as well as translational variation. Because when the user puts his/her palm on the scanner, the angle and position of the palm may change. So, different images are acquired by the scanner according to the input each time. But in our paper we have suggested a rotationalas well as translational-invariant scheme by which the above problem can be overcome while preprocessing the image before the feature extraction of the palmprint. With several palmprint images, we tested our proposed preprocessing system and the experimental results found good.
3 citations
01 Jan 1987
3 citations
••
09 Jul 2011TL;DR: A novel and robust on-line signature verification approach using Hidden Semi-Markov Model (HSMM), which builds a unique HSMM for each identity and estimates the signature baseline in corresponding to the features.
Abstract: Handwritten signature has been extensively adopted as biometric for identity verification in daily life, as it is the most widely accepted personal authentication method. Automatic signature recognition technologies can definitely facilitate the verification process. Many research attempts and advances have occurred in this field, automatic signature verification still is a challenging and important issue. This work presents a novel and robust on-line signature verification approach using Hidden Semi-Markov Model (HSMM). The proposed system comprises three stages. First, dynamic features are extracted according to the local statistical information of velocity, acceleration, azimuth, altitude, and pressure. Next, the extracted features are normalized into unified observation length. To improve the verification accuracy, features with slight variation are clustered into the same class using K-means classification algorithm. Furthermore, the Forward-Backward algorithm is utilized to accelerate the computation of HSMM parameters. Finally, the system builds a unique HSMM for each identity and estimates the signature baseline in corresponding to the features. To assess the recognition performance of the proposed algorithm, experiments were conducted using SVC2004 signature database. Analytical results reveal that the proposed method is very promising.
3 citations