In the design of thermal protection for aero engines and hot-end components such as engine blades, accurate measurement of surface temperature fields is critically important. Domestic mainstream multispectral algorithms have enabled temperature measurement in complex thermal environments, preventing the influence of internal engine background radiation on temperature measurement. However, traditional multispectral algorithms are computationally intensive and require SVD decomposition of spectral data to calculate temperature field data. Additionally, the accuracy of turbine blade positioning signals depends on simulated speed signals, which can result in positional deviations in temperature data. Without speed offset correction, the calculated temperature data cannot accurately reflect combustion temperature non-uniformity or design defects in the disks and blades. This paper innovatively proposes two core optimization strategies: first, the 'multispectral + monochromatic temperature auxiliary correction' strategy, which increases computational efficiency by over 30% compared to existing traditional multispectral optimization algorithms without compromising measurement accuracy; second, an adaptive speed offset correction algorithm that enables dynamic adaptive adjustment of filter parameters, improving offset correction accuracy by 15% compared to existing adaptive correction algorithms. It can handle complex multi-speed operating conditions and addresses the limitations of current temperature measurement algorithms in engineering applications.