The multiangular silhouette photographs (Sections 2.2.2 and 2.2.5) were thresholded to yield black-and-white images by applying an implementation of the Otsu's method [6] in Matlab software to the blue channel of the photographs. The thresholding was preceded by manual drawing of a polygon (in each photograph) that contained only the tree and the white background. The processing chain was the same for both with and without foliage, except that in the photographs without foliage some areas were erroneously detected as tree by the Otsu's method and needed to be removed manually. This was because the contrast between the tree and the background was not always perfect. The resulting black-and-white images were converted into silhouette areas (Stree, [m2]) by multiplying the number of pixels belonging to a tree with the pixel size at the distance of the tree (0.06377 mm2). The pixel size was obtained utilizing the camera's intrinsic (focal length and pixel size) and extrinsic (position and orientation) parameters. These parameters were obtained using photographs taken of a black-and-white checkerboard target in each view angle, and processing with Computer Vision Toolbox™ in Matlab. The processing comprised of performing camera calibration utilizing all photographs, and then solving the position and orientation of the camera in each view (and illumination) angle separately.

Note: The content above has been extracted from a research article, so it may not display correctly.

Please log in to submit your questions online.
Your question will be posted on the Bio-101 website. We will send your questions to the authors of this protocol and Bio-protocol community members who are experienced with this method. you will be informed using the email address associated with your Bio-protocol account.

We use cookies on this site to enhance your user experience. By using our website, you are agreeing to allow the storage of cookies on your computer.