The estimation of canopy attributes from digital cover photography by two different image analysis methods

Proximal sensing methods using digital photography have gained wide acceptance for describing and quantifying canopy properties. Digital hemispherical photography (DHP) is the most widely used photographic technique for canopy description. However, the main drawbacks of DHP have been the tedious and time-consuming image processing required and the sensitivity of the results to the image analysis methods. Recently, an alternative approach using vertical photography has been proposed, namely, digital cover photography (DCP). The method captures detailed vertical canopy gaps and performs canopy analysis by dividing gap fractions into large between-crown gaps and small within-crown gaps. Although DCP is a rapid, simple and readily available method, the processing steps involved in gap fraction analysis have a large subjective component by default. In this contribution, we propose an alternative simple, more objective and easily implemented procedure to perform gap fraction analysis of DCP images. We compared the performance of the two image analysis methods in dense deciduous forests. Leaf area index (LAI) estimates from the two image analysis methods were compared with reference LAI measurements obtained through the use of litter traps to measure leaf fall. Both methods provided accurate estimates of the total gap fraction and, thus, accurate estimates of the LAI. The new proposed procedure is recommended for dense canopies because the subjective classification of large gaps is most error-prone in stands with dense canopy cover.