Font Size: a A A

Establishment Of Morphological Index Model For Cut Stems Based On Image Processing And Uniform Evaluation Method For Morphology Of Cut Stems

Posted on:2019-05-29Degree:MasterType:Thesis
Country:ChinaCandidate:C HeFull Text:PDF
GTID:2348330548461629Subject:Food Science and Engineering
Abstract/Summary:PDF Full Text Request
In order to study the morphology of stems and evaluate the uniformity of morphology of cut stems,the captured images of cut stem were processed with image analysis software,polynomial model of morphology of cut stem fited through morphological index?K?of three kinds of width in different length and model correction by Matlab,by dividing the model of region combinating morphological appearance of cut stem and morphological index to establish a morphological index of cut stem model to analyse the morphology of cut stems.To evaluate the uniformity of morphology of cut stems,fractal dimension and morphological index were calculated by slit island method through measuring perimeter and area of single morphological cut stems.Taking morphology of cut tobacco as the standard,the uniform coefficient model and evaluation method of morphology of cut stems were established and the analysis verified through taking the difference of fractal dimension and morphological index of the cut stem and cut tobacco as the indexes.The main conclusions are as follows:?1?Stem digital images were obtained through CCD camera and processed based on gray histogram by Matlab.The method of basing on Otsu threshold segmentation could accurately extract the stem and ensure the accuracy of image measurement results.?2?Establishment of an index model of morphology of cut stems?K=f?x,y?=0.226+0.090*x-0.100*y-0.001*x2-0.041*x*y+0.042*y2+0.002*x2*y+0.008*x*y2-0.007*y3?,fitting equation root mean square errors?RMSE=0.008?was smaller and determination coefficients reached 0.995,the correction model accuracy higher than before.The morphology of cut stem of?A,?A region?0.039<K<0.169?of the model was the shape of broken filaments;?B,?B,?B?0.051<K<0.399?regional morphology were filamentous cut stem;?C?0.095<K<0.255?regional morphology was filament-like cut stems;?C?0.255?K<0.577?regional morphology was flake-like cut stem;?D and?D?0.255?K<0.772?regional morphology were flake cut stem.The morphological patterns of the six cut stem samples were distinguished by using the morphological model,regional division of the model can accurately reflect the morphology of the stems,and realizing the quantitative and qualitative analysis of the morphology of the cut stems.?3?The morphology of cut stems possessed fractal characteristics and the fractal dimension of cut stems had obvious difference.With the increasing of fractal dimension,the morphology of the cut stems becomed more and more irregular.Determinant coefficient of fitting equation of single morphological cut stems all reached 0.850,meanwhile mean square error and residual error were both smaller,which indicated that the slit island method can accurately calculate its fractal dimension.?4?A method for evaluating uniformity of morphology of cut stems was established preliminarily,taking uniform coefficient as an index,uniform coefficient model of morphology of cut stems was proved accurate and reliable(Ye=4.825X-?-1.149X?-??+0.320),which determinant coefficient reached 0.948 and P test and F test all reached the significant level.When the uniform coefficient of cut stem was less than 0.491,morphology of cut stem was uniform;When the uniform coefficient of cut stem was between 0.491 and 0.813,morphology of cut stem was relatively uniform;When the uniform coefficient of cut stem was between 0.813 and 1.189,morphology of cut stem was ununiform;When the uniform coefficient of cut stem was between1.189 and 1.565,morphology of cut stem was poor uniform;When the uniform coefficient of cut stem was greater than 1.565,morphology of cut stem was poorer uniform.The qualitative analysis of uniform coefficient of morphology of cut stems can be realized by means of quantitative indexes.
Keywords/Search Tags:Morphology of cut stem, Morphological index, Uniform coefficient, Fractal dimension, Image processing
PDF Full Text Request
Related items