Abstract: To address the issue of low imaging quality in disposable medical endoscopes due to hardware cost constraints, which limits their suitability for precision surgical navigation, this study aims to enhance the visibility of hemoglobin-related information by transplanting and optimizing an efficient image enhancement algorithm while ensuring real-time performance. Based on the TI TMS320C6678 DSP platform, the system transplantation of an adaptive hemoglobin enhancement algorithm was completed. Leveraging the parallel computing and memory characteristics of the DSP architecture, deep multi-level optimizations were performed at both the algorithmic and assembly levels: at the algorithmic level, computational complexity was substantially reduced through mathematical reconstruction, look-up tables, and the replacement of time-consuming functions; at the assembly level, execution density was significantly improved by adopting software pipeline structures and instruction reordering techniques. Experimental results indicate that after these systematic optimizations, the execution time of the enhancement algorithm on a single core of the TMS320C6678 was reduced from over 700 milliseconds to under 3 milliseconds, meeting the strict real-time requirements for endoscopic surgical navigation while maintaining effective image enhancement. In conclusion, this optimization scheme enables the efficient and real-time operation of the adaptive hemoglobin enhancement algorithm in disposable endoscope systems, providing a feasible technical solution for improving endoscopic imaging quality on low-cost hardware and demonstrating significant clinical application potential.