Enhancing Road Safety with Real-Time Pothole Detection: A Comparative Study of Watershed Transformation and Seam Carving Algorithms
摘要
The watershed transformation and seam carving algorithms are the main subject of this study’s through evaluation of pothole image segmentation techniques’ efficiency in improving road safety. An analysis between two groups was carried out, one using the seam carving algorithm and the other the watershed transformation. 720 dataset samples were analyzed in total, 360 of which were used for testing and 360 for training. The findings showed that, with an accuracy of 85.29% as opposed to 79.09%, the watershed transformation performed noticeably better than the seam carving algorithm. The results of the statistical analysis showed a 95% confidence interval, a significance level of p = 0.052%, and a G-power value of 80%. In addition, the statistical analysis shows the optimized performance by using watershed algorithm in various applications and thus provide the high accuracy value for pothole image segmentation. Because of its exceptional accuracy the watershed transformation Given its superior accuracy, the watershed transformation method represents a significant advancement over the seam carving algorithm for pothole image segmentation, making it a vital tool for improving road safety systems.