Gesture Recognition using IR sensor camera

In the dynamic field of pattern recognition, there's a notable shift towards human skeleton recognition, particularly in computer vision. Recognizing skeletal gestures presents a challenging task, especially with homogeneous backgrounds adding complexity. Human posture recognition is a crucial aspect within gesture recognition. The rise of low-cost devices like Kinect, along with its Software Development Kit (SDK), offers an opportunity to tackle issues conventional cameras face. This paper explores using Kinect's human skeleton data for posture recognition in a monitoring framework. It investigates six depth distances, each linked to six gesture types derived from human skeletal movements. Results show that Kinect, powered by neural networks, accurately detects six key postures: Start, Stop, Forward, Right, Left, and Normal. This emphasizes the transformative role of neural networks in HCI, revolutionizing gesture recognition for innovative and intuitive interfaces