An Improved SSD and Its Application in Train Bolt Detection

A multi-scale fusion and multi-window target detection method is proposed. Firstly, the model and principle of the classical single shot multi box detector (SSD) method are expounded. From the principle analysis, explain the reason why the SSD method is not good in detecting small targets. On this basis, a multi-window and multi-scale fusion model is proposed and explains its model and working principle. This paper evaluates the detection accuracy of the original SSD model and the improved SSD model by using the same data set. The experimental results show that the accuracy of the original model is about 0.8, and the improvement is about 0.95. Therefore, the improved SSD model is more advantageous and accurate than the traditional SSD detection model in bolt detection.