Objective To solve the problem that the quality of cleaning and disinfection of respiratory endoscope depends on manual visual inspection and lacks objective standards, this research developed an intelligent detection system based on optical imaging and established corresponding quality control standards. Methods An endoscope disinfection optical detection system integrating highresolution endoscope camera, multi-spectral light source and deep learning (DL) algorithm was constructed. From January 2025 to June 2025, 60 BF-Q170 type respiratory endoscopes were included in the respiratory endoscopy center of The Seventh People’s Hospital of Luohe. They were randomly divided into the experimental group and the control group, with 30 units in each group. After all endoscopes were fully and automatically disinfected according to the standard procedures, the experimental group was evaluated by intelligent system, and the control group was evaluated by manual visual inspection. The detection rates of the two groups of pollutants, the average time spent on each endoscopic examination, the consistency among assessors, and the qualified rate of microbial testing were compared. Results The detection rates of protein, blood residue and biofilm in the experimental group were higher than those in the control group (all P＜0.05). Moreover, the average time spent on each endoscopic examination in the experimental group was significantly shorter than that of the control group (P＜0.001). The consistency of the two assessors in the control group was only at a moderate level, while the results of the intelligent detection system were stable. In addition, the qualified rate of microbial testing in the experimental group was significantly higher than that in the control group (P＝0.045). Conclusion This intelligent detection system can achieve rapid, objective and highly sensitive assessment of the cleaning and disinfection quality of respiratory endoscopes. It automatically completes the determination of qualified samples and the interception of unqualified samples. Its overall performance is superior to that of traditional manual visual inspection. The quality control standard established by this system forms an effective closed loop of quality control, which provides technical support for standardization and digital management of endoscopic disinfection.