In order to solve the problem that the traditional vibe algorithm has "ghost" in the early stage of moving target detection and its adaptability to complex environment is not strong, an improved vibe algorithm was proposed. In the background modeling and initialization stage, the "ghost" is eliminated by combining image morphological processing and blending Gaussian background model, and the adaptive update radius and probability are introduced in the background update stage to improve the accuracy of the algorithm to adapt to multi-objective complex environment. The experimental results show that on the premise of ensuring a certain real-time performance, the improved algorithm can quickly and effectively eliminate the phenomenon of "ghost", and has a strong adaptability to the complex environment, which provides an important reference for the real-time detection of moving targets.