|
Description:
|
Technologies currently used for cotton contaminant assessment suffer from some fundamental limitations . These limitations severely restrict the ability of existing technologies to accurately detect and classify contaminants in cotton . Such inaccuracies result in the misassessment of the cotton quality , and have a serious impact on its economic value .
The fundamental limitations of existing methods include the inability to detect contaminants under the surface of cotton , the inability to accurately measure shapes and sizes , sample preparation requirements , and poor spatial resolution . These limitations may be easily overcome by the use of x -ray tomographic imaging , which allows for highly accurate imaging of the internal features of an object in a non -destructive fashion .
This thesis describes in detail the design of a GUI based interactive cotton contaminant analysis tool . Through the use of an x -ray microtomographic scanner and image processing algorithms , it is shown that x -ray tomographic imaging can provide very accurate information regarding shape , size , and density of cotton contaminants . This information has been analyzed using a fuzzy -logic -based classification scheme to create a highly accurate contaminant analysis tool .
Despite its obvious advantages , x -ray imaging does have some drawbacks , principle among which pertains to the time taken to perform the procedure . These drawbacks , along with possible solutions have also been discussed in this thesis . It is our firm belief , however , that if realized in real -time , this procedure will have a definite impact on the cotton cleaning process , and indeed on the economic value of cotton . |