Operating rooms are the most expensive and critical medical resources in hospitals, and scheduling efficiency directly impacts patient waiting times, healthcare workforce burden, and overall hospital operational efficiency. However, most hospitals still manage surgical scheduling using paper or basic electronic forms, relying on manual interpretation and adjustments, which are not only time-consuming and labor-intensive but also prone to resource wastage and communication errors due to information inconsistencies or omissions. This study aims to develop a "Hospital Surgery Scheduling Optimization System" that integrates image and document recognition technology with scheduling algorithms. It automatically extracts room numbers, times, doctors, and surgical information from surgical scheduling photos, converts them into structured data, and proposes optimized operating room schedules while adhering to clinical and management constraints. The system provides a web interface for users to upload files, view results, and compare pre- and post-optimization schedules, ultimately outputting standardized scheduling reports to enhance operating room utilization and scheduling transparency, thereby achieving the goals of smart healthcare and digital process transformation. Preliminary results show that the system's OCR module achieves a field recognition accuracy rate of over 85% under medium-quality images, significantly reducing the manual input time by approximately 70%. The scheduling optimization algorithm improved operating room utilization from the original 65% to 85%, reduced daily overtime hours by 20-30%, and decreased surgical delay cases by 15% in simulated scenarios. Additionally, the system integrates a carbon emission estimation model, predicting a 10-15% reduction in the operating room carbon footprint post-optimization, demonstrating its potential for efficiency and sustainable healthcare. Future efforts will involve clinical validation to confirm the long-term benefits.