With the recent development of immersive interaction systems, a variety of applications based on gesture recognition that enable intuitive interaction are being developed. The most important aspect of gesture recognition technology is a high gesture recognition rate, which requires that the gestures be defined and learned in advance. However, the process of re-editing data depending on the learning results is troublesome in most gesture recognition and learning systems. In this study, we propose a system that makes re-editing easier by visualizing the learning result according to the gesture learning results. Moreover, the proposed system enables more complex and precise gesture recognition by merging data from multi-sensors and processing them as one gesture. Experimental results from the system implementation show that the editing of visualized learning data was very intuitive and the average gesture recognition rate was 95%.