The present paper describes the use of Digital Image Processing and Internet of Things for gestures recognition using depth sensors available on Kinect R device. Using the open source libraries OpenCV and libfreenect, image data are translated and used for communicating Raspberry Pi embedded Linux system with PIC microcontroller board for peripheral devices controlling. LED is triggered according to the hand gesture representing the corresponding number. Data are stored in a PHP Apache server running locally on Raspberry Pi. The proposed system can be used as a multifunctional tool in areas such as learning process, post-traumatic rehabilitation and visual and motor cognition time. Using image binarization and Naive-Bayes classifier, the achieved results show error lower than 5%.