Abstract
With the rapid development of high-speed railways, any objects intruding railway clearance will do great threat to railway operations. Accurate and effective intrusion detection is very important. An original Single Shot multibox Detector (SSD) can be used to detect intruding objects except small ones. In this paper, high-level features are deconvolved to low-level and fused with original low-level features to enhance their semantic information. By this way, the mean average precision (mAP) of the improved SSD algorithm is increased. In order to decrease the parameters of the improved SSD network, the L1 norm of convolution kernel is used to prune the network. Under this criterion, both the model size and calculation load are greatly reduced within the permitted precision loss. Experiments show that the mAP of our method on PASCAL VOC public dataset and our railway datasets have increased by 2.52% and 4.74% respectively, when compared to the original SSD. With our method, the elapsed time of each frame is only 31 ms on GeForce GTX1060.
Highlights
As of 2018, the length of high-speed railway lines in China had exceeded 29,000 km
Experiments show that the mean average precision (mAP) of our method on PASCAL VOC public dataset and our railway datasets have increased by 2.52% and 4.74% respectively, when compared to the original Single Shot multibox Detector (SSD)
The mAP of our method on PASCAL VOC and railway datasets increased by 2.52% and 4.74%
Summary
As of 2018, the length of high-speed railway lines in China had exceeded 29,000 km. With the increasing of speed, any object intruding railway clearance will pose a major threat to high-speed trains. Effective intrusion detection methods are of great significance to the safety of a high-speed railway operation. The disturbance of strong environment light and severe weather will lead to false alarms To solve this problem, the convolutional neural network (CNN) [2] was used to detect objects intruding railway clearance. Liu [12] extracted bounding boxes from CNN multi-layer feature maps to improve the speed and accuracy with the SSD model. In order to solve this problem, this paper proposes an improved SSD method for railway clearance intrusion detection. The deconvolution structure is introduced into the SSD network to improve the detection ability of small objects. A recursive SSD network pruning method is proposed to reduce the model parameters and calculation load with the convolution kernel L1 norm criterion within a 1% precision loss.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
