Implementation and Development of Static Hand gesture recognition in a real time using Diagonal Sum algorithm for Gestural Language

This manuscript deems the proposal over utilization of computer digitized vision over the gesture recognition. Gesture language is a language that determines the requirement over combining the finger gesture, its orientation, arm and hand movement, facial & body expression that simultaneously explores and advertises the people thoughts. The digital camera makes the recording of live motion streams of pictures with which the acquisition of image is made with the assistance of interface. The training of system is made over each sort of Figureureureureure gestures as representing (5,4,3,2 or 1) atleast in a single time. Later on, the test symbol is delivered and the system makes a try for detecting it. In this proposed study the detection of Figureureureureure gestures is made using the strategy of image processing. The system makes the detection of cumulative finger count. Later on, it makes the identification of individual fingers above the palm. During the processing, it initially makes the detection of skin tone (Color) from the acquired image by the utilization of filter. The image is allowed to process through subsequent steps in order to depict the correct count of fingers. The model makes the detection over the nearer point from the threshold value. The detection of image is made as per the centroid value. Later on, the implication of certain steps is made for enhancing the normal image to an efficient image so that the exposure of fingers is made. Finally, the model makes the detection and decides the finger count and advertises the calculation to the tester. As a result, the classification is done using artificial neural networks based on the previously formulated training model that has been built and realized with more than 92.5% of accuracy in the finger gesture recognition. In this study, the comparison has also been enumerated with other state of art algorithms designed by many researchers. The classification has been illustrated with diagonal sum algorithm and in the phase of progression, this DSA algorithm makes the removal of background noises and finally the conversion of image is made into the binary image and also realized statistically. This infers clearly for differentiating various finger gestures very accurately with its evaluation metrics.