The Particle Swarm Optimization (PSO) algorithm has long been applied to magnetic positioning. However, due to its limitations, many studies have explored hybrid strategies to optimize the performance of positioning systems. This paper combines the Greedy algorithm (GA) with the Branch and Bound (B&B) algorithm, proposing an improved hybrid strategy for the GA. The performance of the classic GA and the improved GA is compared in terms of positioning accuracy and computational efficiency. Experimental results indicate that, compared to the classic GA, the improved GA significantly enhances positioning accuracy and computational efficiency, successfully meeting the precision requirements for intestinal capsules. Specifically, the improved hybrid strategy markedly reduces positioning errors, achieving high precision standards. This study introduces a novel approach to magnetic localization in the medical field, particularly for the precise positioning of intestinal capsule endoscopes. The experimental results demonstrate that the proposed improved GA effectively enhances positioning accuracy, providing strong support for future research and applications.