Open AccessAn Efficient Gesture Recognition with ABC-ANN Classification and Key-Point Features Extraction for Hand Images
Author(s) -
Satinderdeep Kaur,
Er. Nidhi Bhatla
Publication year - 2019
Publication title -
international journal of innovative technology and exploring engineering
Language(s) - English
Resource type - Journals
ISSN - 2278-3075
DOI - 10.35940/ijitee.j1153.0881019
Subject(s) - gesture , computer science , gesture recognition , interface (matter) , scale invariant feature transform , artificial intelligence , key (lock) , feature extraction , support vector machine , computer vision , orientation (vector space) , boosting (machine learning) , point (geometry) , user interface , pattern recognition (psychology) , geometry , computer security , mathematics , bubble , maximum bubble pressure method , parallel computing , operating system
With the advent of an electronic device in humanization, there has been various usage of easing human-computer interface. However, novel methods and techniques have been developed for boosting the HCI interface. In the world of technology, including gestures in the human-computer interface, become a crucial exploration field. The gesture has become a communication approach with the computer interface and users. Hand Gesture works as a natural boundary for serving as an inspiring strength for investigating gesture classifications. The hand gesture is used in a wide variety of applications like as recognition technology, controlling of system, computer system, electronic device, home appliances, flying equipment controlling, and so forth. The hand gesture is described as the production of the motion or the gestures by hands; it determines the expression through a signature pattern that leads to the interface between the system and individuals. However, hand gesture plays the main role in the establishment of the novel human and system interface. In this research, a classification technique with ABC –ANN is developed to enhance the detection rate and the SIFT algorithm has been implemented to extract features of hand images. Features are described as - i) Assignment (ii) Localization (iii) Orientation and (iv) SIFT key-points. The extraction of features is done through the key point format. Along with that, classification and selection of the specific features of gesture image acquired based on categories. Experimental analysis is done based on parameter metrics (testing rate, validation rate, recognition rate) to improve the detection rate.