This research aims to clarify the relationship between image properties of aerial photograph and structural factors of Castanopsis cuspidata coppice woodland and Chamaechiparis obutusa artificial forest, and to develop an algorithm of tree crown analysis using the images. The aerial photographs were taken in May, 1999, the season of flowering Castanopsis cuspidata, and converted to monochrome 45cm ground resolution digital images. Fourteen plots were chosen in the field to measure vegetation and forest structure. Correlations were analyzed between image properties such as SD (standard deviation) and CF (cumulative frequency of tree crown edge) and forest structural factors such as tree crown area, tree density, relative illuminance under tree crown, etc. As a result of only Chamaechiparis obutusa artificial forest, a -0.79 correlation between SD and CF was obtained. Also, a 0.79 correlation between SD and the relative illuminance under tree crowns was found. Furthermore, in terms of CF, a 0.72 (tree crown area) and a -0.82 (tree density) correlation was found. The results suggested that these images contained the information of the tree crown.