Comparison of dimension reduction and optimization techniques in integration of spatial and spectral information of satellite images to improve change detection in urban areas
سال انتشار: 1394
نوع سند: مقاله کنفرانسی
زبان: انگلیسی
مشاهده: 400
فایل این مقاله در 13 صفحه با فرمت PDF قابل دریافت می باشد
- صدور گواهی نمایه سازی
- من نویسنده این مقاله هستم
استخراج به نرم افزارهای پژوهشی:
شناسه ملی سند علمی:
RCEAUD01_051
تاریخ نمایه سازی: 12 تیر 1395
چکیده مقاله:
In the recent decades, cities had unplanned growth and rapid expansion towards agricultural lands. Application of remotely sensed data and image processing techniques is a good method for change detection in urban areas. This is appropriate for responsible and thoughtful urban management. Addition to the techniques for change detection, input data to the algorithm have special influence in the accuracy of the results. To do this, spatial and spectral properties of the remote sensing multi-temporal images have been used. They have been integrated by Genetic Algorithm optimization and dimension reduction technique of Principal Component Analysis. Ten states have been examined to apply these techniques. Two approaches of comparison after classification and direct classification have been used to produce change map of urban areas of Tabriz from 1989 to 2010 using two temporal images of TM5. By Using, evaluation data, the accuracy of change map and runtime of the algorithm have been examined. The results have indicated that the best results are obtained for the state that in which all the original spatial and spectral properties of images constitute search space of genetic optimization without any dimension reduction method. Kappa coefficient for the produced change map is equal to 88 percent. This is about 11 percent better than classification with original bands and 7 percent better than dimension reduction method. However, the runtime of GA method is longer than that of dimension reduction. As the accuracy of the results is preferred to run time, the GA method is quite advised.
کلیدواژه ها:
نویسندگان
A Rastegar
instructor of Department of engineering and technology, Golestan University, I.R of Iran
مراجع و منابع این مقاله:
لیست زیر مراجع و منابع استفاده شده در این مقاله را نمایش می دهد. این مراجع به صورت کاملا ماشینی و بر اساس هوش مصنوعی استخراج شده اند و لذا ممکن است دارای اشکالاتی باشند که به مرور زمان دقت استخراج این محتوا افزایش می یابد. مراجعی که مقالات مربوط به آنها در سیویلیکا نمایه شده و پیدا شده اند، به خود مقاله لینک شده اند :