Abstract: To promote the assisted parking (AP) technology, accurate and visual actual scene information needs to be provided to divers. Due to the complexity of the automotive environment and the variety of interference information in the automotive millimeter wave (MMW) radar SAR images, it is difficult to achieve target detection by target separation directly. Therefore, a target detection method is proposed for automotive MMW radar SAR image based on MSER. This method is based on a priori information of the size of the automotive after imaging, and provides drivers with parking space usage by detecting the MSER of the SAR image, dividing and marking vacant parking area. Morphological filtering is used to smooth the boundary of MSER regions to optimize the target detection results. At the same time, the accuracy of target detection is endured by marking the target with an adaptive width rectangle box. Finally, the effectiveness of the proposed method is verified by simulation and analysis of the measured data.