The Study And Application Of The Tree-Ring Micro Density Based On The Virtual Instrument And Image Processing Technology

Density is an important index for evaluation of the characteristics and quality of wood and the methods for its measurement have developed from the average density to the tree-ring micro-density with the development of the Virtual Instrument and Image Processing,Firstly,the Microdensitometer and tree-ring image method were analyzed systemically.Resort to the virtual instrument and X-ray image processing technology,the dissertation submitted three systems,which included Microdensitometer 3CS Data Acquisition System(3CSDAS),Tree-Ring Image Analysis System (TRIAS),Tree-Ring Analysis System(TRAS),and put forward the method of Digital Radiography image direct measuring tree-ring micro-density for the first time.And then,these systems were applied to detecting the main components and variation of poplar I-69(Populus deltoides CV.I-69/55)tree-ring width and tree-ring density,climatic chronological table and climatic factors response of pinus taiwanensis's as well.The main conclusions of this paper were as follow:1.Digital development of Microdensitometer 3CS based on virtual instrument technology(1)The analog signals of recording stylus of Microdensitometer 3CS were translated into digital signals through the virtual instrument technology and the PC-7422 data acquisition card;The Visual C++ were employed as the development platform for Virtual Instrument,which includes data input,data acquisition,graphics amendment,density calibration,data saving,and printing module.The application of the system could improve the Microdensitometer accuracy and efficiency.(2)The result of Microdensitometer 3CS measurement was optical density,which had been calibrated.Calibration results showed that optical density could be applied to measure the wood density, the thickness and uniformity of objects completely.Compared to the measuring results of OFTEX Microdensitometer,the average relative errors of digital Microdensitometer 3CS for measurement of width and tree-ring micro-density were less than 3%.The results of measurement of paper thickness using digital Microdensitometer 3CS showed that the variation coefficient of relative thickness was 38.63%in the rice paper,while 7.69%in the copy paper.The initial broken position of paper sheet was on relatively thin sides.2.Using X-ray image directly for measurement of tree-ring micro density(1)X-ray image is the total projection that X-ray penetrated different densities and thickness of objects.X-ray images can eliminate the heartwood-sapwood tonal variation,the sample cranny,resin canal spots,polish scratches and other errors.With adaptive filtering,the histogram equalization techniques,the mathematical morphology and the theory of image difference,the noise could be avoided,the image brightness could be tone up and the background errors could be rectified.(2)TRIAS,which use mixed programming technology of Visual C++ and MATLAB,includs images reading and writing,image processing,density calibration,data saving,the charts display,printing and other functions.TRIAS could be used to measure both the tree-ring width and the tree-ring density,and date visually at the same time.It is not necessary to use the special densitometer but X-ray photographic devices only when measuring the wood density by TRIAS,which provides a new approach for the development of micro density measurement by TEIAS.(3)The X-ray image measuring results were gray scale,which were needed for calibration. Calibration results showed that image gray scale could be applied to measure the wood density,the thickness and uniformity of objects completely.The comparison test results showed that:for the species which springwood and late wood were difficult to identify,TRIAS results were superior to that of Microdensitometer.Compared to the Microdensitometer results,the average relative errors of tree-ring width and tree-ring micro density measured by X-ray image method and DR were less than 3%,5%. Which meant the measurement of tree-ring width and tree-ring micro density were feasible.3.Design of Tree-Ring Analysis System(TRAS)based on LabVIEW(1)TRAS,which employed the LabVIEW as the development plafform,included 4 functional modules:data acquisition,data processing,data storage and result display.TEAS could analyze not only the data measured by densimeter,but also the data measured by image method.TRAS could delete disturbance ring and inserts lost ring by the method of combination automatic judgment with man-machine interaction;demarcate boundary points between springwood and late wood automatically by the method of dynamic threshold method and gradient method;read the tree-ring counts,the constituents of tree-ring width and density directly by mathematical model.(2)Compared to CRED system and ARA system,TRAS had higher accuracy in tree-ring boundary precision,but inferior in demarcating boundary between spfingwood and latewood.Comparatively,the error of gradient method was superior in poplar I-69,while error of dynamic threshold method was superior in Chinese fir.4.Studies on main component and its variation of poplar I-69 tree-ring width and tree-ring micro density(1)The poplar I-69 was chosen as the sample,the dates of the components of tree-ring width and density and analysis of the radial and axial variation of the components were gained by TRIAS and TRAS.The results suggested that:the variation of ring width and density of poplar I-69 was consistent in various height;the micro density was lower at heartwood of poplar,and higher near bark,the variogram was S shape on the whole;the variation of the component of poplar I-69 ring density was little,and the ring width was large.(2)The Radial variation of the density of poplar I-69 increased along with the tree age.The maximum density fluctuated heavy,the minimum density decreased along with increase of the tree age on the whole.The density of the earlywood and latewood increased along with the tree age.The radial variation of every indicator of the tree-ring components was significant within individual trunk, otherwise no difference among trunks.The components of the tree-ring density increased along with the tree height,there into,the tree-ring average density,earlywood density and minimum density increased a little,while the tree-ring maximum and latewood density increased much.Comparatively,the radial variation of the density oftbe poplar I-69 was larger than that of the axial variation.(3)The correlation analysis of the components of the tree-ring showed that the tree-ring density and the tree-ring width increased along with the tree age,the tree-ring density and the tree-ring width had positive correlation,which indicated the possibility of selecting poplar I-69 clones combined growth rate and quality. 5.Establishment of the tree-ring components chronologies and its response to climatic factors of ptnus taiwanensis(1)The Pinus aiwanensis was chosen as the sample,took pictures by Digital X-ray Radiography device,acquired the spectra of tree-ring micro density and the data of tree-ring width and micro density components by TRIAS and TRAS.The statistical results showed that the susceptibility of tree-ring width was 0.216 and tree-ring micro density was 0.120,and the variation of tree-ring width and tree-ring micro density was 6.59%and 3.53%respectively.(2)Huangshan in Anhui province is the semi-humid to humid areas,so the variation of Pinus aiwanensis tree-ring is not significant and it's difficult to demarcate years.This dissertation first put forth the method of demarcating years by tree-ring image and micro density spectrum sections.It could decrease global errors to demarcate years within sections,and improve accuracy and reliability by analyzing the tree-ring variation within sections.A new approach of dendrochronology by Digital Radiography images was explored.(3)The relation between the components of tree-ring and the climate change showed that the tree-ring width and tree-ring micro density of Pinus aiwanensis in high altitude humid area had positive correlation with sunshine hours and the annual average temperature,no correlation or negative correlation with amount of precipitation and negative correlation with the frost days.By the correlation function significance it's relevant between the density and temperature.