پديد آورنده :
مهرعليان، سهيل
عنوان :
ارائه روشي نوين مبتني بر تحليل مولفه هاي اصلي براي شناسايي انسان در تصوير
مقطع تحصيلي :
كارشناسي ارشد
گرايش تحصيلي :
هوش مصنوعي و رباتيك
محل تحصيل :
اصفهان: دانشگاه صنعتي اصفهان، دانشكده برق و كامپيوتر
صفحه شمار :
سيزده،96ص.: مصور،جدول،نمودار
يادداشت :
ص.ع.به فارسي و انگليسي
استاد راهنما :
مازيار پالهنگ
توصيفگر ها :
بينايي كامپيوتر , تشخيص الگو , قطعه بندي تصاوير هوايي
تاريخ نمايه سازي :
4/8/92
استاد داور :
محمدرضا احمدزاده،عبدالرضا ميرزايي
دانشكده :
مهندسي برق و كامپيوتر
چكيده فارسي :
به فارسي و انگليسي: قابل رويت در نسخه ديجيتالي
چكيده انگليسي :
A New PCA Based Method for Pedestrian Detection in Images Soheil Mehralian s mehralian@ec iut ac ir Date of Submission 2013 01 23 Department of Electrical and Computer Engineering Isfahan University of Technology Isfahan 84156 83111 Iran Degree M Sc Language FarsiSupervisor Maziar Palhang palhang@cc iut ac irAbstract One of the important goals of researchers in Artificial Intelligence and Robotics is to have a machine canlike a human On the way to achieve this goal the machine should have a good perception of theenvironment One of the essential information that a machine my have about its environment is to know whoare where and what they are doing Various solutions have been proposed to answer these questions thatalmost all of them are in the realm of Computer Vision which shows its importance in this application In thisresearch we have proposed a new method for pedestrian detection in images and videos Our method usessliding windows to search through images Each window is divided into overlapping cells from whichfeatures are extracted The feature that we extracted to describe each window is based on the analysis of thegradient distribution of each cell After gradient distribution of a cell is computed the PCA is applied on itand a mathematical expression is calculated as a feature of that cell Then features are classified using SVM Finally the learnt model was tested on MIT and INRIA pedestrian datasets The results show that our methodis comparable with other methods and is more robust than others against noisy images In addition topedestrian detection the proposed method was used to segment aerial images into urban and non urbanregions Results of this application also show the method can segment images with high accuracy and speed Keywords Computer Vision Pedestrian Detection Pattern Recognition Aerial ImagesSegmentation
استاد راهنما :
مازيار پالهنگ
استاد داور :
محمدرضا احمدزاده،عبدالرضا ميرزايي
