A.V. Krevetsky, Y.A. Ipatov

Monitoring forest resources are widely used visual inventory methods. Despite the high efficiency of methods based on analysis of remote sensing data, they have limitations for a variety of tasks, restored from ground-based measurement methods. Visual approach to the estimation geometric parameters of the stand, are inherent features of forming a sensor installed at a given altitude in the azimuth plane. Images trunks are relatively uniform brightness areas with a nearly vertical orientation. Based on the analysis of such images is necessary to define the boundaries and the width of the band-pass objects to selection objects with a width of not less than the specified. Shooting space ranging scanner runs from the same position and in the same direction as the digital video camera. The obtained algorithms are tested in real conditions on the data of a digital camera and a laser scanner. Sharing optical and ranging sensor and using the received algorithm processing provides new quality of the taxation results: the resolution of visible parts by overlapping objects, estimate their width at different distances, context of automatic correction of results, automatic determination or appoint the size of investigated territory. The testing results of software implementation for proposed methods confirm the models adequacy of the observed images and signals, as well as the effectiveness of automated measurements. The proposed software system can be used when carrying out engineering and scientific research in the field of forestry to improve their reliability and performance.