Automatic identification for standing tree limb pruning

Abstract

To meet the demand of automatic pruning machines, this paper presents a new method for dynamic automatic identification of standing tree limbs and capture of the digital images of Platycladus orientalis. Methods of computer vision, image processing and wavelet analysis technology were used to compress, filter, segment, abate noise and capture the outline of the picture. We then present the arithmetic for dynamic automatic identification of standing tree limbs, extracting basic growth characteristics of the standing trees such as the form, size, degree of bending and their relative spatial position. We use pattern recognition technology to confirm the proportionate relationship matching the database and thus achieve the goal of dynamic automatic identification of standing tree limbs.