Gait analysis can objectively assess abnormal walking, and some walking parameters can help recognize the disease. Existing commercial systems are either too expensive and require attachments to the body or have limitations in detecting abnormal gait. A vision system has been proposed to address this. However it had limitations where the accuracy was inferior in some parameters such as gait phase, step length and width, etc. Therefore we developed a Tactile sensor-based treadmill to detect gait phase, step length, and width. A pilot test was performed and analyzed through an infrared marker-based motion capture system to compare the accuracy of the proposed system. The measured spatiotemporal gait parameters were analyzed through mean and standard deviation and compared to the baseline system. As a result of the experiments, it was confirmed that higher step width performance was achieved compared to previous studies. Future studies will validate the system with many participants and conduct clinical studies on gait recognition through abnormal gait analysis.