Hand Gesture Recognition: An Approach of Multiple
Object Tracking
Shubha Chaturvedi
1
and Dr. Manoj Kumar Rawat
2
1-2
Sage University /Computer Science & Engineering Department, Indore, India
Email: shubha.dubey07@gmail.com, profrawat.sage@gmail.com
Abstract—In the recent year, Object Detection and Tracking technology is one of the
challenging and most emerging topic in Computer Vision. Multiple Object Tracking &
Detection (MOTD) has gained a lot of importance in various fields such as Autonomous driving,
Monitoring security, Surveillance, Health-care monitoring and Gesture recognition etc. Gesture
recognition is widely used in various fields of intelligent driving, Virtual reality, and Human-
computer interaction. With the development in various technologies such as Deep Learning,
Artificial Intelligence and Human-Computer interaction a new revolution arise in field of
computer vision. We endeavor to provide a thorough review on the development of this gesture
recognition approach in recent decades. Here our major discussions are about the benefits and
limitations of existing approach where focus is on the method of feature extraction in
spatiotemporal structure .Various research difficulties that could be pursued will be the
research directions.
Index Terms— Object Detection, Gesture recognition, CNN, YOLO.
I. INTRODUCTION
In this era of innovation, recent advances have done in the research area of Human computer interaction (HCI)
and gesture recognition is the rapidly developing area to provide the communication between a human and
computer. Human gesture is the simplest way of non-verbal communication, humans provide input in various
ways such as facial expression, and various body movements of head, hand and body but apart from face the
most widely used communication is hand gesture movement. Hand Gesture may be classified as static and
dynamic where static gesture is the representation of the stable hand whereas dynamic gesture is the series of the
movement and further dynamic may be categorized as isolated or continuous.
The first and foremost step in the hand gesture recognition is the Hand Gesture attainment which may be further
analyzed in the two forms such as the sensor based and vision based. Further, sensor based approach focuses on
the use of the instruments or the sensors physically attached with the hand which observes the movement of the
hand, figures and its trajectory. Some sensor based technology which was analyzed in the past few years are
glove –based approach, Electromyography, mechanical and electromagnetic techniques. Vision-based approach
captures the data from the cameras, LIDAR, RADAR etc. and some other methods also exist.
II. HCI (HUMAN COMPUTER INTERACTION) SYSTEM
HCI System may be classified according to their input and output such as Uni-modal and Multimodal. Various
Grenze ID: 01.GIJET.9.1.322_1
© Grenze Scientific Society, 2023
Grenze International Journal of Engineering and Technology, Jan Issue