Research and implementation of intelligent parking lot path planning algorithm based on breadth-first search

Parking lot path planning is a key research direction in urban intelligent traffic management, with the core objective of rapidly planning the optimal path from the starting position to the target parking spot. The traditional Breadth-First Search (BFS) algorithm has limitations in practical applications due to redundant node expansion and insufficient real-time performance, which restricts the efficiency and effectiveness of path planning. This paper combines OpenCV technology and Haar-cascade classifiers to implement vehicle parking detection, dynamically generate the logical map of the parking lot, and proposes an optimized algorithm to address the shortcomings of BFS. The optimization methods include a pruning strategy to reduce redundant path expansions, a path caching mechanism to improve the efficiency of repeated queries, and a priority expansion strategy to enhance goal-directedness. Experimental results show that the optimized algorithm effectively reduces redundant node expansions, improves adaptability to dynamic environments, and significantly reduces search time and memory consumption, providing an efficient solution for path navigation and dynamic resource allocation in intelligent parking lots.