Thesis (MScFor)--Stellenbosch University, 2013. / ENGLISH ABSTRACT: Information on the external shape and internal properties of a tree such as the branch structure, tree ring widths and formation, and defects such as resin pockets is important for many forest –and wood science researchers as well as for the forest and wood processing industries. Resin pockets are internal defects associated with some softwood species, and are undesirable in some wood applications such as furniture and veneer. A tool that is often used in research to obtain this information is the Computed Tomography scanner. The high cost of the scanner, as well as the cost of transporting logs to the scanner, limits the application range of this equipment. An alternative, lower cost method would be beneficial for many researchers interested in tree shape and macroscopic internal tree characteristics. The objectives of this study were to: - Design, construct and test a mobile system that can be used in field to obtain a three‐dimensional model of a log or tree stem indicating selected macroscopic internal characteristics and, - Interpret results to explain the occurrence and causes of resin pockets in Pinus patula from the Mpumalanga escarpment, South Africa. A system for dissecting and measuring trees in field was designed, constructed and tested. The metal frame was used to cut a stationary, fixed log into discs and obtain digital images of crosssections along the stem axis. The frame was capable of processing logs with maximum dimensions of 3 m in length and 45 cm in diameter. Software was developed to convert the digital images into three‐dimensional models of logs and trees to demonstrate properties such as the external shape of the trees, branch structure and pith location. Properties such as ring width were measured from the images obtained. The study was conducted in the Mpumalanga escarpment to obtain three‐dimensional models of Pinus patula trees and to establish the reason for formation of resin pockets in these trees. Four 3 m logs from twenty‐four trees from three compartments were dissected and digitally reconstructed into three‐dimensional models. A total of 61 cross sections with resin pockets were identified from the 2750 cross cut images taken 20 cm apart along the longitudinal axis of the log. It was not possible to accept or reject conclusively any of the existing hypotheses for the cause of resin pocket formation. The formation of Type 2 resin pockets was most probably due to felling damage during thinning operations. It seems likely that wind damage and possibly insects might be responsible for the formation of Type 1 resin pockets. / AFRIKAANSE OPSOMMING: Kennis in verband met die interne eienskappe van ‘n boom soos takeienskappe, jaarringwydte en vorm, en harsholtes is belangrik vir sommige bos ‐en houtkundige navorsers. Die toerusting wat huidiglik algemeen gebruik word om inligting van hierdie aard te bekom is ‘n CT skandeerder. Die metode is egter duur as gevolg van hoë kapitale koste van die toerusting sowel as die vervoerkoste van stompe na die CT‐fasiliteit. ‘n Meer ekonomiese alternatief sal vir baie navorsers van waarde wees. Harsholtes in hout lei tot afgradering van hout en fineer. Die doel van die studie was twee‐sydig: - Om ‘n stelsel te ontwerp en bou wat ‘n drie‐dimensionele model van ‘n boom kan skep wat sommige makroskopiese, interne eienskappe weergee, en - Om die voorkoms en oorsaak van harsholtes in Pinus Patula van die Mpumalanga platorand te ondersoek. ‘n Raam is ontwerp en vervaardig wat in die plantasie gebruik word. Die raam kan stompe van tot drie meter lengte met ‘n maksimale diameter van 45 cm hanteer. Die raam is gebruik om snitte deur die stompe te maak, ‘n afstandmeting tot by die gesaagde oppervlak word deur ‘n lasermeter geneem en ‘n kamera neem ‘n foto. Sagteware word dan gebruik vir die uitkenning van sekere dele in elke beeld sowel as die samestelling van ‘n drie‐dimensionele model. Data‐insameling met die raam is in Mpumalanga gedoen. Die raam is in die tydperk getoets en het goed gefunksioneer. Vier stompe elk vanaf vier‐en‐twintig bome vanuit drie afsonderlike kompartemente is verwerk op die raam en volledige drie‐dimensionele modelle geskep vir hierdie bome insluitend hul eksterne vorm, takeienskappe en pitvorm (“pith location”). Ongeveer 2750 foto’s van dwarssnitte is tydens data‐insameling geneem. Net 61 van die foto’s het harsholtes bevat. Geen definitiewe oorsaak van harsholtes kon bepaal word nie. Tipe 2 harsholtes word waarskynlik as gevolg van skade tydens uitdunning van bome gevorm. Wind en insekte lyk na die mees waarskynlike oorsake van Tipe 1 harsholtes.