The method for suppressing OH fluorescence images background must be researched to improve the accuracy of image processing and velocity calculation in the complex combustion Hydroxyl Tagging Velocimetry(HTV) measurement. Firstly, the background interference in the complex combustion is analyzed and identified, and the noised signal simulation model is constructed. Secondly, an adaptive difference method combined with adaptive spatial filtering is proposed for background suppression image preprocessing, based on the characteristic of residual background and OH tagging line signal. Interference in the intensive combustion reaction area is eliminated through the adaptive difference method which is utilized by inducing spatial transformation. The background interference elimination results background is optimized through adaptive spatial filtering difference method by exploiting the spatial distribution difference between the signal and the residual background. Finally, simulation and experiments are carried out to test the effectiveness of this method. Results show that, with this image preprocessing method, the signal to noise ratio is raised by 11.83 dB, the peak signal to noise ratio is elevated by 8.66 dB, while the velocimetry calculation error is improved to 1.2% demonstrating that this method meets the accuracy requirements of HTV measurements and is ready for practical application.