The invention relates to a product appearance detection method based on cloud edge collaborative model optimization and an implementation system thereof. The detection method comprises the following steps: S1, establishing a basic data set; s2, training a YOLOv3-tiny model and a YOLOv3 model, and respectively deploying the YOLOv3-tiny model and the YOLOv3 model at an edge server and a cloud platform; s3, detecting and recognizing, by the YOLOv3-tiny model, the picture, when the picture is detected to be qualified, sending the picture and a detection resultto the cloud platform, and executing the S5, if the detection result is unqualified, determining that the result is suspected to be unqualified, sending the result to the cloud platform, and performing S4; s4, performing secondary detection on the suspected unqualified pictures on the cloud platform; if the secondary detection is not qualified, outputting a result, and ending the process, and if the secondary detection is qualified, performing S5; and S5, storing the qualified picture in the cloud platform, outputting a result, and ending. According to the method, the problems of accuracy, flexibility, time delay and data utilization rate are solved by using a working mode of cloud edge collaboration.