پديد آورندگان :
قلي نژاد، سعيد دانشگاه صنعتي خواجه نصيرالدين طوسي - دانشكده مهندسي نقشه برداري , شاد، روزبه دانشگاه فردوسي مشهد - دانشكده فني و مهندسي - گروه مهندسي عمران , صدوقي يزدي، هادي دانشگاه فردوسي مشهد - دانشكده فني و مهندسي , قائمي، مرجان دانشگاه فردوسي مشهد - دانشكده فني و مهندسي
كليدواژه :
تجزيهي طيفي , تصاوير فراطيفي , فاكتورگيري ماتريس نامنفي (NMF) , يادگيري گروهي
چكيده فارسي :
به تازگي روشهاي تجزيهي طيفي تصاوير فراطيفي، بهعنوان ابزاري قدرتمند در شناسايي عوارض موجود در پيكسلهاي مختلط، بهطور گستردهاي مورد اقبال پژوهشگران قرار گرفتهاند. از ميان الگوريتمهاي ارائهشده براي تجزيهي طيفي تصاوير فراطيفي، فاكتورگيري ماتريس نامنفي (NMF) به علت اعمال قيد نامنفي بودن بر فراوانيهاي حاصل از تجزيهي طيفي و همچنين استخراج همزمان طيف و فراواني اعضاي خالص، بيش از ساير روشها مورد توجه قرار گرفته است. عليرغم اين تواناييها، NMF به علت داشتن تابع هدف نامحدب داراي جوابهاي محلي فراواني است كه در مطالعات مختلف با افزودن قيدهايي به تابع هزينهي آن، تلاشهايي براي دستيابي به نتايج بهينهي سراسري صورت پذيرفته است. با اينحال، روشهاي بر مبناي NMF همچنان داراي جوابهاي محلي هستند. در پژوهش حاضر با استفاده از يك روش تكراري و با تكيه بر تئوري يادگيري گروهي و تركيب وزندار نتايج بهدستآمده از تكرارهاي مختلف الگوريتم تجزيهي طيفي L1/2-NMF، فرآيند استخراج طيفها و فراوانيهاي حاصل از اين الگوريتم بهبود يافته است. روش پيشنهادي روشي غيرپارامتريك و از نظر رياضي روشن است كه ميتوان فرآيند پيشنهادي در آن را به الگوريتمهاي پيشرفته تري از تجزيه ي طيفي تعميم داد. روش پيشنهادي بر روي دادههاي مختلف مصنوعي و واقعي اجرا گرديده است. نتايج حاصل از آزمايشهاي موجود در اين پژوهش، بر روي هر دو دسته از دادههاي فراطيفي، حاكي از كارايي اين روش نسبت به روشهاي مشهور در شناسايي عوارض موجود در پيكسلهاي مختلط است.
چكيده لاتين :
Images material identification plays an important role in remotely sensed image processing and its applications. Hyperspectral images, which contain a lot of narrow spectral bands of the electromagnetic spectrum, have a great potential for information extraction from remotely sensed images and material identification. Due to the low spatial resolution of hyperspectral cameras, material mixing, and light multi scattering, these images usually contain a lot of mixed pixels which face material identification with many problems. Recently, hyperspectral unmixing methods have been widely considered by the researchers as a powerful tool for identifying materials in the mixed pixels. Some of the algorithms, proposed for hyperspectral unmixing, are based on the linear mixture model (LMM) and the others are based on nonlinear mixture model (NLMM). LMM-based algorithms are more simple and commonly used methods for hyperspectral unmixing. Among various algorithms, based on LMM, non-negative matrix factorization (NMF) has attracted the most attention due to essentially implying non-negativity of the endmembers and their corresponding abundances, and moreover, simultaneously extracting spectral signature and abundances of the endmembers. In spite of these capabilities, NMF leads to local minima due to its non-convex objective function. In this regards, various studies have attempted to lead NMF results to the global optimum by imposing some additional constraint to the main objective function of NMF. However, NMF-based methods still suffer from the problem of falling into local minima. To tackle this problem, an iterative post-processing procedure, based on an ensemble learning technique, has been presented in this paper. The main goal of this paper is to demonstrate the ability of ensemble learning to improve the hyperspectral unmixing results in a simple and non-parametric manner. To this end, NMF with sparse constraint is performed in several iterations, and then, the results of each of these iterations are weightened on the basis of identifying a primary endmember that certainly exists in the image. Weightening is done with calculating spectral angle distance (SAD) metric between the true and extracted spectral signatures of the primary endmember. Usually, there is prior information about the hyperspectral images such as some existed materials or the number of materials in the images. Therefore, it is always possible to find a primary endmember in an image. The accuracy of identifying primary endmember is extended the accuracy of identifying other endmembers of their corresponding abundances. Final mixing and abundances matrices are determined using weighted combinations of the mixing and abundances matrices, extracted from each of the iterations. The proposed procedure is nonparametric and mathematically clear which can be extended to more advanced algorithms of hyperspectral unmixing.