The invention relates to an object detection method, in particular to an improved YOLOv2 object detection method in a cluttered environment. A YOLOv2 object detection model is used to detect objects,the length and width of the detection object boundary box are output, and the output length and width are reduced by K1 and K2 times, wherein K1=1. 47 and K2=1. 0612. As the length and the width of the boundary frame of the detection object output from the YOLOv2 model are respectively reduced by K1 and K2 times, each boundary frame contains only one object after reduction, and the intersection ofthe boundary frames of the plurality of object is reduced, so that all the objects in the picture can be detected, the three-dimensional coordinate information of the target object and the error of the distance between the target object and the camera are greatly reduced, and the detection accuracy is improved.