Abstract: Favored foliage colors in autumn are genetically determined by the genes of tree species,different seasonal colors and transition are theorically controlled by the pigments (chlorophyll,anthocyanidin,xanthophyll etc.) in leaves,and the transitions of geographical environment and seasons drive and model them in time and space.In 2014,digital lapse photography was used to collect the phenological image time series of Ginkgo biloba growing in Dujiangyan Campus,then ENVI programming was used to calculate temporal changes of three vegetation indices,NDVI,Gcc and ExG.A computer program,TIMESAT,was applied to fit the double logistic function of a time series data of NDVI,Gcc and ExG of Ginkgo biloba in Dujiangyan.The maximum value method of the half curve was used to calculate the key phenological date (the serial days of a year) of G.biloba:the start date of growing season(SOS),the end date of growing season(EOS),the maximum of growing and the length of season.Analysis was made of decline percentage in fall along the curves of NDVI,Gcc and ExG respectively,and the results showed the percentage (the serial days of a year) between 50% and 90% had a good response to the released time of four red leaf tourim guide index in the same year of Sichuan Province,especially DOY from the NDVI curve was almost the same.The result showed the phenological monitoring method by digital camera-based remote sensing was a potential tool for forest tourism research.