With the purpose of evaluating the development degree of preferential flow under dry and wet farmland in Three Gorges Reservoir Area, the statistical sample variability analysis method and morphological image processing technology were used to analyze variability of the preferential flow areas in the field scale soil profile dyed images. The results showed that soil wetting front infiltration rate and soil dyeing image parameters, including wetting front infiltration rate, dye coverage, uniform infiltration depth, preferential flow fraction and length index, changed in the characterization development degree of preferential flow wet farmlands were higher than those of dry farmlands. The preferential flow evaluation method was proposed according to the comparison with soil dyeing image parameters. The same result showed the accuracy of preferential flow area dyed image variability analysis method to analyze the development degree of preferential flow. The new method is simpler than the others, and can improve the accuracy of the visual evaluation dyed images. It is also better to combine statistics and morphology to quantitatively analyze the development degree of soil preferential flow.