The designed method introduces a novel approach for road humps detection and distance estimation to promote the safety of intelligent transportation systems (ITS) under low-light conditions and congested traffic. The key contribution of this work is to integrate CLAHE-based contrast enhancement and a lightweight MobileNet backbone into a faster R-CNN architecture, which improves feature extraction efficiency without sacrificing accuracy. After the hump is detected, the distance to the hump is estimated using stereo vision, which computes disparity and uses triangulation to determine real-world depth. Unlike existing methods, the proposed approach combines road hump detection with distance estimation using CLAHE for low-light image enhancement, allowing for more accurate and dependable perception in difficult lighting situations. In contrast to traditional techniques, our method allows for the accurate recognition of both marked and unmarked humps in a variety of road and lighting conditions, including background clutter and occlusions. Experiments were conducted on an expanded Mendeley dataset and real-world driving scenarios. The proposed model outperformed the baseline faster R-CNN and YOLO variations with a 94% detection accuracy. The overall root mean square error (RMSE) for the distance estimation was 0.62 m. The major advantage of this system is the improvement of transport management, autonomous vehicles, and especially for safe driving.

In an advanced intelligent transportation system vehicle recognition and classi f ication is very significant. In current research trend, recognition of vehicles is done byusingmachinelearning (ML)andcomputervisiontechniques. Vehicle’s multi-view images or videos with different lighting conditions are annotated and given to the deep neural network to build an automated system to recognize the vehicles models. The augmentation of data can increase the number of sam ples in learning, with the small available datasets. Geometric transformations, brightness changes, and different filter operations are applied to the data through data augmentation. Furthermore, be orthogonal experiments we determine the optimal data augmentation method to obtain 96% accuracy in results. Detailed information is reported based on the classification of four different types of vehi cles and the results show that convolutional neural network with 16 layers deep techniques are effective in solving challenging tasks while recognizing moving vehicles.