Abstract: Although, voice is main way used for communication, facial expressions and body language are also used while interacting with others. In our regular life, hand gestures are used to communicates with deaf and mute people by sign language, to express the feelings like ‘stop’ or ‘bye’. With the advancement in machine learning, artificial intelligence and applications like Human Computer Interaction (HCI), hand gestures are getting widely used to interact with machines and computers. This study presents the effective approach to identify dynamic hand gesture. Hand region detection is a vital task in dynamic hand gesture recognition module. For that purpose an improved method is introduced by combining HSV, YCgCr and YCbCr color spaces. After detection of hand portion from the video sequence their color texture and edge features are extracted. Then for gesture recognition an artificial neural network is used. Thus, this proposed system can be viewed as a complete dynamic hand gesture recognition system that can be used for various purposes.