Evaluation of the carbon, water, and energy balances in evergreen coniferous forests requires accurate in situ and satellite data regarding their spatio-temporal dynamics. Daily digital camera images can be used to determine the relationships among phenology, gross primary productivity (GPP), and meteorological parameters, and to ground-truth satellite observations. In this study, we examine the relationship between seasonal variations in camera-based canopy surface indices and eddy-covariance-based GPP derived from field studies in an Alaskan open canopy black spruce forest and in a Japanese closed canopy cedar forest. The ratio of the green digital number to the total digital number, hue, and GPP showed a bell-shaped seasonal profile at both sites. Canopy surface images for the black spruce forest and cedar forest mainly detected seasonal changes in vegetation on the floor of the forest and in the tree canopy, respectively. In contrast, the seasonal cycles of the ratios of the red and blue digital numbers to the total digital numbers differed between the two sites, possibly due to differences in forest structure and leaf color. These results suggest that forest structural characteristics, such as canopy openness and seasonal forest-floor changes, should be considered during continuous observations of phenology in evergreen coniferous forests.