The invention provides a soil image brightness controllable enhancement method based on weighted Gaussian subtraction fitting, and the method comprises the following steps: S1, collecting a soil image, and calculating the gray scale probability density Porg (y) of a Y component of the soil image; s2, constructing a soil image Y component histogram based on the Y component gray level and the gray level probability density of the soil image; s3, determining the definition domain of the Y component of the soil image as [ya, yb], and performing weighted Gaussian subtraction fitting based on the Y component histogram of the soil image to obtain a weighted Gaussian subtraction fitting curve of the Y component of the soil image; s4, performing brightness controllable enhancement processing on the basis of the weighted Gaussian subtraction fitting curve of the soil image Y component, and performing RGB conversion on the enhanced soil image to obtain an enhanced RGB soil image; through brightness controllable enhancement, soil images collected under inconsistent natural illumination conditions are adjusted, so that the soil images are very similar to real soil images collected under certain specific illumination conditions, and the method is used for machine vision soil type identification, thereby reducing the influence of different natural illumination conditions on soil image identification, and improving the subsequent soil type identification precision.