The detection of rail surface defects is vital for high-speed rail maintenance and management. The CNN-based computer vision approach has been proved to be a strong detection tool widely used in various industrial scenarios. However, the CNN-based detection models are diverse from each other in performance, and most of them require sufficient training samples to achieve high detection performance. Selecting an appropriate model and tuning it with insufficient annotated rail defect images is time-consuming and tedious. To overcome this challenge, motivated by ensemble learning that uses multiple learning algorithms to obtain better predictive performance, we develop an ensemble framework for industrialized rail defect detection. We apply multiple backbone networks individually to obtain features, and mix them in a binary format to obtain better and more diverse sub-networks. Image augmentation and feature augmentation operations are randomly applied to further make the model more diverse. A shared feature pyramid network is adopted to reduce model parameters as well as computation cost. Experimental results substantiate that the approach outperforms single detecting architecture in our specified rail defect task. On the collected dataset with 8 defect classes, our algorithm achieves 7.4% higher mAP.5 compared with YOLOv5 and 2.8% higher mAP.5 compared with Faster R-CNN.

Spent fuel reprocessing is of great significance to the nuclear fuel cycle and the sustainable development of nuclear energy. At the same time, nuclear security radiation incidents in the spent fuel reprocessing plant are also related to national personal and property safety, which play a pivotal role. In this paper, the spent fuel reprocessing plant is divided into four plant areas: the main process area, the three-waste area, the auxiliary equipment area, and the pre-plant area, which are further subdivided into 12 evaluation units. The expert scoring method is used to score and evaluate the possibility of eight basic nuclear accident types in each area, namely radioactive dispersal device, computer nuclear security, destruction of nuclear facilities, transportation nuclear security, internal threat, potential threat, illegal transfer, and theft. According to the professional titles, length of service, education and other qualifications of experts, different weights are assigned to the experts. The scoring results are applied to the Fault Tree Analysis (FTA) of nuclear security events as the probability of basic events, so as to obtain the risk of each basic event. At the same time, the fuzzy comprehensive evaluation method and probability–mathematical statistics method are used to evaluate each evaluation unit to determine the risk of each evaluation unit and the plant area. There results show that the main process area has the highest risk degree, while the pre-plant area has the lowest risk degree, and there is a 1.5-fold relationship. This research provides theoretical and technical support for the safety management and operation of spent fuel reprocessing plants. The analysis results of this paper can be used as a reference for the proportion of nuclear security protection improvements in each plant area, so as to achieve an efficient safety protection effect. The research method in this paper can be also applicable to other similar places by providing as input the corresponding probability of occurrence to obtain the index of its risk degree, so as to reasonably allocate funds and manpower and reduce risks.

Image-based rail defect detection could be conceptually defined as an object detection task in computer vision. However, unlike academic object detection tasks, this practical industrial application suffers from two unique challenges, including object ambiguity and insufficient annotations. To overcome these challenges, we introduce the pixel-wise attention mechanism to fully exploit features of annotated defects, and develop a feature augmentation framework to tackle the defect detection problem. The pixel-wise attention is conducted through a learnable pixel-level similarity between input and support features to obtain augmented features. These augmented features contain co-existing information from input images and multi-class support defects. The final output features are augmented and refined by support features, thus endowing the model to distinguish between ambiguous defect patterns based on insufficient annotated samples. Experiments on the rail defect dataset demonstrate that feature augmentation can help balance the sensitivity and robustness of the model. On our collected dataset with eight defected classes, our algorithm achieves 11.32% higher mAP@.5 compared with original YOLOv5 and 4.27% higher mAP@.5 compared with Faster R-CNN.