Deep Convolutional Neural Networks for Pavement Crack Detection using an Inexpensive Global Shutter RGB-D Sensor and ARM-Based Single-Board Computer

Abstract

Pavement distress assessment is a significant aspect of pavement management. Automated pavement crack detection is a challenging task that has been researched for decades in response to complicated pavement conditions. Current pavement condition assessment procedures are extensively time consuming, expensive, and labor-intensive. The primary goal of this paper is to develop a cost-effective and reliable platform using a red, green, blue, depth (RGB-D) sensor and deep learning detection models for automated pavement crack detection on a single-board ARM-based computer. To the best of our knowledge, for the first time, a pavement crack data set is prepared using a global shutter RGB-D sensor mounted on a car and annotated according to the Pascal visual object classes protocol, named PAVDIS2020. The proposed data set comprises 2,085 pavement crack images that are captured in a wide variety of weather and illuminance conditions with 5,587 instances of pavement cracks included in these images. A unified implementation of the Faster region-based convolutional neural networks and single shot multibox detector meta-architecture-based models is implemented to evaluate the accuracy, speed, and memory usage trade-off by using various convolutional neural networks-based backbones and various other training parameters on PAVDIS2020. The proposed pavement crack detection model was able to classify the cracks with 97.6% accuracy on PAVDIS2020 data set. The detection model is able to locate pavement crack patterns at the speed of 12 frames per second on a passively cooled Raspberry Pi 4 single-board computer.

Deep CNN Approach with Visual Features for Real-Time Pavement Crack Detection

“…Initial efforts in automated pavement crack detection relied on basic digital imaging, utilizing simple edge-detection algorithms within 2D images, as documented in [15]. While groundbreaking at the time, these methods grappled with considerable constraints, including low detection accuracy, vulnerability to varying environmental conditions, and an inability to process complex real-world data effectively [16].…”

HE-YOLOv5s: Efficient Road Defect Detection Network

“…Liu et al [ 15 ] proposed a two-step road defect detection and method based on convolutional neural networks, first by using an improved YOLOv3 model to detect the dataset and an improved U-Net to segment the cracks in the dataset for training, and finally achieved an F1 score of 90.58% on the CFD dataset. Pouria Asadi et al [ 16 ] developed an efficient and beneficial platform that combines speed and detection performance, using RGB-D sensors and target detection models such as Faster-RCNN and SSD, achieved 97.6% performance for road defect detection on the PA VDSI2020 dataset.…”