Development of an Infrared-Based Sensor for Finger Movement Detection

Abstract

With the increasing interest in smart devices and convenient remote control, the need for accurate wireless means of control has become imperative. This gives rise to the growing research interests in the area of gesture and finger movement detection. In this paper, a suitable design exploring some techniques involved in hand and finger movement detection, using depth-sensing infrared cameras embedded on Xbox Kinect Module is presented. First, 3-D images are generated and filtered along the z-axis, then two distinct techniques; Haar-Like Features, and Deep Learning using a Convolution Neural Network (CNN), are performed on the images to detect hands movement. In order to evaluate the robustness of the proposed technique, useful metrics like, Precision, Recall, F1-Score and Accuracy are used to evaluate the efficiency of the technique. Results show that while the deep learning model showed the most accurate results with a weighted accuracy of 1.0 (due to the absence of noise in the images), a weighted accuracy of 0.97 is observed for the Haar-Like features. Finally, the Haar-like features technique appears to run faster due to its static nature whereas, the deep learning model is quite slow in terms of running time. Overall, these findings point to the conclusion that the deep learning model is a better technique for detecting hands movements despite its longer running time.