<![CDATA[Motion in sports workouts plays a crucial role in evaluation and learning. This design aims to develop a sports workout movement recognition device based on image processing with high accuracy. The design methodology involves acquiring image data of workout movements and training a recognition model using Mediapipe, which is a deep learning-based object detection algorithm. Image data of workout movements is obtained from push-up and sit-up exercises performed in front of the Sports Movement Acquisition module. The training process of the Mediapipe model utilizes this input data to recognize movements such as push-ups and sit-ups. Initial experimental results indicate that the system can recognize workout movements with satisfactory accuracy. Furthermore, this research includes testing the system with varying distances to determine its accuracy at different ranges. In further experiments, the system can be refined or improved to achieve higher accuracy. This research contributes significantly to the development of sports workout movement recognition technology using the Mediapipe algorithm, focusing on the classification of push-up and sit-up movements. The system can be used in training, evaluation, and assessment of specific workout movements, namely push-ups and sit-ups. The integration of image processing and deep learning in this field holds potential for further development in movement analysis and training for both athletes and beginners. By combining modern technology and sports science, this research opens new opportunities for understanding and enhancing workout movements, particularly in push-up and sit-up exercises, potentially providing significant benefits in athlete training and the development of workout sports.]]>
