X-ray tube is the most commonly used equipment in X-ray fluorescence spectrum analysis as excitation source whose primary spectrum has become a major source of the background of X-ray fluorescence spectrum. Background subtraction should go before further analysis of this spectrum. The accuracy of the estimation of the background directly affect the results of subsequent processing steps. In this paper, the tube excitation component of X fluorescence spectrum is analyzed with an estimation method of background intensity against their background characteristics, and structuring theoretical spectral lines based on the measured spectral lines in order to evaluate the effect of spectral processing algorithms. The method utilizes the measured X-ray fluorescence spectrum which does not contain the characteristic peak of the spectrum, In order to estimate the composition of the original spectrum of the X-ray tube. It uses the spectrum that contains the continuous background for interpolating the entire measured spectrum, thereby avoiding overlapping the spectral characteristic peaks or improperly estimating full width at half maximum. This paper compared SNIP, Fourier transform method and the background subtraction method using the measured spectral lines. Using this method to estimate the background is closer to the theoretical background. The results showe that the method for X-ray fluorescence spectrum by X-ray tube excitation is accurate in this article, this method can be used to deduct continuous background, and good applied effectiveness has been achieved in the background subtraction of the actually measured X-ray fluorescence spectrometry.