پديد آورنده :
دانشور، الهه
عنوان :
كاهش رنگ تصاوير ديجيتالي تهيه شده از مناظر طبيعي
مقطع تحصيلي :
كارشناسي ارشد
گرايش تحصيلي :
شيمي نساجي و علوم رنگ
محل تحصيل :
اصفهان: دانشگاه صنعتي اصفهان، دانشكده نساجي
صفحه شمار :
ده، 93ص.: مصور، جدول، نمودار
يادداشت :
ص.ع. به فارسي و انگليسي
استاد راهنما :
حسين ايزدان
استاد مشاور :
داريوش سمناني
توصيفگر ها :
استتار ديجيتالي , توصيف رنگي , نايكنواختي روشنايي
تاريخ نمايه سازي :
94/2/8
استاد داور :
عبدالكريم حسيني، هاله خليلي
تاريخ ورود اطلاعات :
1396/09/26
چكيده انگليسي :
Colour reduction of digital images obtained from natural scenes Elaheh Daneshvar e daneshvar@tx iut ac ir Date of Submission 11 Jan 2015 Department of Textile Engineering Isfahan University of Technology Isfahan 84156 83111 Iran Degree M Sc Language PersianSupervisor Dr H Izadan izadan@cc iut ac irAbstractCamouflage has long been one of the major factors to be considered in military and passive defense and ithas been used in military applications in order to protect the soldiers and their equipment in modern wars Digital camouflage as the latest version of the camouflage pattern has attracted the military industry sattention Camouflage in visible Vis region of the electromagnetic spectrum is one of the camouflagemethods and its aim is to hide the camouflaged object from the vision of the human or devices worked invisible region The purpose of this present research is the development of the patterns with optimized colorsin Vis region by the application of the color reduction algorithm and a characterized digital camera Thetarget area for the extraction of the pattern was considered the jungle area of Isfahan Therefore manyimages were captured from the jungle areas of the Isfahan and four of them were considered as therepresentation of them The digital camera was characterized by a multiple regression and an alternatingleast squares techniques using a custom made color chart and a color checker DC Also the digital camerawas characterized by the srgb2lab function of the image processing toolbox of the MATLAB program Then the representative colors of each of the images were extracted in CIELAB color space and by theapplication of the combined Kohonen neural network and k means clustering techniques A referencedigital pattern was also considered and its image was captured by the digital camera and the representativecolors were allocated to the area of this reference pattern on the basis of the nearest color By this method four camouflaged patterns were made In order to evaluate the performance of the characterizationtechniques a photo simulation method was applied using a characterized LaCie 320 display In photosimulation technique each of the captured images of Isfahan jungles were considered as the background individually and the camouflaged patterns were randomly displayed on five different predetermined areasof the backgrounds 30 observers whose normal vision was tested by Ishihara method evaluated theperformance of the different characterization techniques on the basis of the detection time It was concludedthat the Kohonen kmeans color reduction technique performed well It was also concluded that the srgb2labcharacterization method was performed excellently However the multiple regression or alternating leastsquares technique using either color checker DC or custom made chart were performed inferior incomparison to the reference pattern Keywords digital camouflage color reduction characterization natural scene nonuniform illumination
استاد راهنما :
حسين ايزدان
استاد مشاور :
داريوش سمناني
استاد داور :
عبدالكريم حسيني، هاله خليلي
