<p>Illegal parking on pedestrian crosswalks disrupts traffic flow and endangers pedestrian safety, particularly in crowded urban areas. Conventional surveillance systems continuously transmit large amounts of image or sensor data, leading to excessive bandwidth and energy consumption. This study introduces a goal-oriented semantic communication framework for intelligent monitoring of pedestrian crosswalk violations. The proposed system employs drones or fixed roadside sensing units that locally process raw data to identify meaningful events—specifically, the presence of vehicles obstructing pedestrian crossings. Only semantically relevant information is transmitted to a central control unit when such events occur, thereby minimizing communication overhead while maintaining timely and accurate detection. By integrating event-triggered sensing, edge intelligence, and goal-oriented communication, the system achieves substantial reductions in data transmission and energy use compared to traditional monitoring approaches. The proposed framework provides an efficient and scalable solution for smart city traffic management and pedestrian safety enforcement.</p>