<p>Recently, self-driving vehicles have been introduced with several automated features, including lane-keep assistance, queuing assistance in traffic jams, parking assistance, and crash avoidance. These autonomous vehicles (AVs) and intelligent visual traffic surveillance systems mainly depend on cameras and sensor fusion systems. One of the most critical issues in developing AVs and driver assistance systems is their poor performance under adverse weather conditions (ADWC), such as rain, snow, fog, and hail. Consequently, it is crucial to develop an effective model to detect objects, such as vehicles and pedestrians, under challenging conditions, including adverse weather. This paper proposes a deep learning (DL) model, specifically a Modified ReLU-based Bidirectional Long Short-Term Memory (MRBLSTM), for autonomous vehicle detection (AVD) in ADWC. It is mainly composed of six stages. Initially, the proposed system collects the data from the publicly available dataset for VD. After that, image restoration is done as a preprocessing step on the collected data using the Retinex algorithm. Next, the Modified Canny Edge Detection Operator (MCO) algorithm preserves edges in the restored images using a bilateral filter and Otsu’s approach. Then, the features are extracted from the edge-detected images using the HoG, HOG-like feature, LBP, and SIFT feature extraction models. The optimal features are selected from the extracted features to lower the dimensions of the features using the Cauchy mutation-included Coyotes Optimization Algorithm (CMCOA). Finally, the vehicles are detected using the MRBLSTM. The proposed method is tested on the DAWN dataset, and experimental results demonstrate its effectiveness, outperforming state-of-the-art VD approaches under ADWC.</p>