Magnetically controlled capsule endoscope (MCCE) is an emerging tool for thediagnosis of gastric diseases with the advantages of comfort, safety, and noanesthesia. In this paper, we develop algorithms to detect and measure humangastric peristalsis (contraction wave) using video sequences acquired by MCCE.We develop a spatial-temporal deep learning algorithm to detect gastriccontraction waves and measure human gastric peristalsis periods. The quality ofMCCE video sequences is prone to camera motion. We design a camera motiondetector (CMD) to process the MCCE video sequences, mitigating the cameramovement during MCCE examination. To the best of our knowledge, we are thefirst to propose computer vision-based solutions to detect and measure humangastric peristalsis. Our methods have great potential in assisting thediagnosis of gastric diseases by evaluating gastric motility.