Abstract: Compared with traditional biological radars, MIMO radars can obtain the azimuth information of human targets, which had obvious advantages. In order to reduce the interference caused by human target motion and near-field environmental clutter, an improved MIMO bio-radar human target detection and tracking joint adaptive algorithm was proposed. First, the method of image morphology was introduced to pre-process the clutter of the radar imaging results. Second, the improved adaptive threshold method was used to detect the target. Finally, the track correlation method was used to obtain the accurate trajectory of the target movement, and the detection threshold was adjusted according to the target position. The measured data shows that the method can effectively filter out clutter interference and has good tracking continuity.