A Novel Approach for Detection of Pavement Crack and Sealed Crack Using Image Processing and Salp Swarm Algorithm Optimized Machine Learning

Abstract

During the phase of periodic survey, sealed crack and crack in asphalt pavement surface should be detected accurately. Moreover, the capability of identifying these two defects can help reduce the false-positive rate for pavement crack detection. Because crack and sealed crack are both line-based defects and may resemble each other in shape, this study puts forward an innovative method based on computer vision for detecting sealed crack and crack. This method is an integration of feature extraction based on image processing and metaheuristic optimized machine learning. Image processing is used to compute features that characterize visual appearance and texture of the pavement image. Subsequently, Salp Swarm Algorithm integrated with multiclass support vector machine is employed for pattern recognition. Based on experimental results, the newly developed method has achieved the most desired predictive performance with an accuracy rate = 91.33% for crack detection and 92.83% for sealed crack detection.

Research on high-precision recognition model for multi-scene asphalt pavement distresses based on deep learning

“…Particularly, mAP@50 provides a comprehensive assessment of the model's performance across the entire dataset. The specific calculation formulas are as follows: (10) where the term True Positive (TP) represents the number of samples correctly classified as distressed, False Positive (FP) represents the number of samples incorrectly classified as distressed, and False Negative (FN) represents the number of distressed samples that were not correctly identified. P(R) represents the Pr-Re curve plotted at different classification thresholds.…”

“…With ongoing advancements in computer vision and richer datasets, machine learning significantly improves road defect detection accuracy and efficiency. Support vector machine (SVM), a potent algorithm in defect image analysis, demonstrates superior classification performance 9,10 . Wang et al 11 adopted the minimum rectangular coverage (MRC) model to group pavement cracks and utilized a two-stage SVM model to classify 10,134 MRCs from four highway segments, achieving an accuracy rate of 88.07%.…”

Detection of Road Cracks Using Convolutional Neural Networks and Threshold Segmentation

“…Image Pre-processing Crack Detection Feature Extraction In order to detect cracks in the pavement, [17] used a linked domain technique (directed segmentation expansion algorithm) to produce a binary image, which was then processed. The threshold segmentation approach is typically used in conjunction with other algorithms to improve segmentation accuracy due to its reliance on gray-level features and its susceptibility to noise.…”

“…Most of the earliest approaches [15,16,17] relied on threshold processing algorithms that assumed the crack pixel was darker than its neighbors. The threshold segmentation approach was frequently utilized in early image segmentation techniques because of its ease of use and speed.…”

Using Image Processing for Automatic Detection of Pavement Surface Distress

“…As a result, this kind of camera digitizes the image from the beginning, even if it is still a representative representation of the light intensity because of the signal amplitude. [13,14] The image captured in this research was taken with Cannon EOS 500D (figure 1). And the digital image is presented in figure 2.…”