ronever / uapd-pavement-distress-dataset Goto Github PK

Abstract: Pavement distress detection is crucial in the decision-making for maintenance planning. Unmanned aerial vehicles (UAVs) are helpful in collecting pavement images. This paper proposes the collection of pavement distress information using a UAV with a high-resolution camera. A UAV platform for pavement image collection was assembled, and the flight settings were studied for optimal image quality. The collected images were processed and annotated for model training. Three state-of-the-art object-detection algorithms—Faster R-CNN, YOLOv3, and YOLOv4, were used to train the dataset, and their prediction performances were compared. A pavement image dataset was established with six types of distress. YOLOv3 demonstrated the best performance of the three algorithms, with a mean average precision (MAP) of 56.6%. The findings of this study assist in the inspection of non-destructive automatic pavement conditions.

Abstract: Automatic pavement distress detection is essential to monitoring and maintaining pavement condition. Currently, many deep learning-based methods have been utilized in pavement distress detection. However, distress segmentation remains as a challenge under complex pavement conditions. In this paper, a novel deep neural network architecture, W-segnet, based on multi-scale feature fusions, is proposed for pixel-wise distress segmentation. The proposed W-segnet concatenates distress location information with distress classification features in two symmetric encoder-decoder structures. Three major types of distresses: crack, pothole, and patch are segmented and the results were discussed. Experimental results show that the proposed W-segnet is robust in various scenarios, achieving a mean pixel accuracy (MPA) of 87.52% and a mean intersection over union (MIoU) of 75.88%. The results demonstrate that W-segnet outperforms other state-of-the-art semantic segmentation models of U-net, SegNet, and PSPNet. Comparison of cost of model training and inference indicates that W-segnet has the largest number of parameters, which needs a slightly longer training time while it does not increase the inference cost. Four public datasets were used to test the generalization ability of the proposed model and the results demonstrate that the W-segnet possesses well segmentation performance.