The increasing use of aerial sensing systems in intelligent transportation and automation environments has enabled the large-scale acquisition of urban imagery for vehicle monitoring and analysis. However, reliable vehicle detection using aerial sensor data remains challenging owing to degraded signal quality and complex scene structures in urban environments. To address these issues, we propose a sensor signal processing and automation framework for robust vehicle detection and classification in aerial imagery. The framework begins with Wiener filtering for noise suppression and image enhancement, followed by Markov random field-based segmentation to improve spatial consistency and object delineation. For feature representation, the EfficientNet deep learning model is employed to extract discriminative features with computational efficiency suitable for automation systems. In addition, a hybrid feature extraction strategy combining Haar wavelet transform-based scale-invariant representations with the histogram of oriented gradients is introduced to capture complementary structural and texture information from aerial sensor signals. The resulting features are classified using the DenseNet-121 model to improve feature propagation and classification performance within an integrated processing pipeline. Experimental results on the Vehicle Aerial Imagery and Roundabout Aerial Images datasets demonstrate the effectiveness of the proposed framework, achieving precision scores of 97.6 and 95.1%, respectively. Overall, the proposed framework provides an effective solution for sensor signal collection, processing, and system integration in intelligent transportation systems and smart city automation applications.