| Abstract |
In decision-making, NS permits the representation of information with three membership functions: indeterminacy (I), false (F), and truth (T). All components in an NS have indeterminacy, non-, and membership degrees that are autonomous and vary from (0-1). This generates NS particularly appropriate in composite decision-making situations where information is incomplete, ambiguous, or contradictory, which allows strong and more complex solutions and analysis. Detecting road damage accurately and quickly enables the capability of road maintenance agencies to generate timely maintenance to road surfaces, retain optimum road conditions, enhance the safety of transportation, and reduce transportation charges. Research on road damage detection using AI models achieved more attention at present, particularly in smart cities. This paper develops a Boosting Road Damage Detection using DEMATEL with Bipolar Neutrosophic Dombi and Siberian Tiger Optimization (BRDD-DBNDSTO) algorithm. The presented BRDD-DBNDSTO technique is mainly intended to improve the accuracy and reliability of road damage classification for intelligent smart city infrastructure. To accomplish this, the BRDD-DBNDSTO technique employs adaptive bilateral filtering (ABF) using image preprocessing to effectively enhance image quality by reducing noise. Then, the SqueezeNet method was used to create a collection of feature vectors. For the classification and detection of road damage, the DEMATEL with bipolar neutrosophic Dombi model is exploited. At last, the Siberian tiger optimization (STO) algorithm is used to adjust the parameters related to the classifier model. To guarantee the improved performance of the BRDD-DBNDSTO method, an extensive experimental study was carried out and the gained outcomes illustrate the improvement of the BRDD-DBNDSTO model across the existing techniques. © 2025, American Scientific Publishing Group (ASPG). All rights reserved. |
