عنوان مقاله [English]
Textures are useful indices for extracting information from aerial photographs. The main objective of this study was to map forest canopy cover density based on aerial photographs using textural indices in the middle part of Zagros forests, west ofIran. Four Arial photographs with scales and spatial resolution of 1:40000 scale and 0.56 meter respectively were used. Photos were orthorectified by Arial camera parameters, digital elevation model, fiducial marks and ground control points. Texture indices including standard deviation, mean and contrast in difference dimension of co-occurrence matrix were extracted from aerial photographs.Forestcanopy cover density classification was done on original and textural bands. Useful index and co-occurrence matrix dimension were chosen using accuracy assessment by maximum likelihood algorithm. Results indicated that classification using texture indices had higher accuracy than original channel. Based on the results of this study, mean index with matrix dimension, 13×13 pixel showed the best accuracy in comparison with other indices. Overall accuracy and kappa coefficient were obtained 61.36% and 0.48, respectively using mean index. Although, the results showed an average accuracy, this method is still useful for mapping forest cover. Finally, black and white aerial photographs can be used to extract more accurate information using texture.