<![CDATA[Motion recognition is a technology that enables a system to recognize movements carried out on human body parts using cameras, sounds and sensors. Parts of the human body that can be used as indicators include cases, hands, feet, heads, etc. In this study a system was built that could classify hand movements using the MPU6050 sensor on wearable devices used on the hands. The K-Nearest Neighbor (KNN) classification method is used in this study because it has a fast data processing speed in doing classification. The test data is obtained through a wearable device that has a device to read the accelerometer sensor data, then sent using the RF24 wireless module to Arduino Uno as a receiving device to enter the movement classification process and display the classification results on the LED conditions. Tests in this study give 100% accuracy in gesture recognition in hand waving movements to turn on the lights, 100% in left hand waving movements to dim the lights, and 100% in hand movements down to turn off the lights. Besides that, the computation time is tested to classify the movement with the KNN method in this study. computational time from testing gives an average time of 94.3 milliseconds on 30 motion tests or 10 tests on 3 types of motion.]]>
Copyrights © 2019
