UNIVERSITI PUTRA MALAYSIA MODELLING OF GEOMETRIC CORRECTION AND RELATIVE RADIOMETRIC NORMALIZATION FOR NEAR EQUATORIAL EARTH OBSERVATION SATELLITE IMAGES HAYDER ABD AL-RAZZAQ ABD FK 2015 86 M P MODELLING OF GEOMETRIC CORRECTION AND RELATIVE U RADIOMETRIC NORMALIZATION FOR NEAR EQUATORIAL EARTH OBSERVATION SATELLITE IMAGES T H G IBy R HAYDER ABD AL-RAZZAQ ABD Y P O C © Thesis Submitted to the School of Graduate Studies, Universiti Putra Malaysia, in Fulfillment of the Requirements for the Degree of Doctor of Philosophy October 2015 All material contained within the thesis, including without limitation text, logos, icons, photographs and all other artwork, is copyright material of Universiti Putra Malaysia unless otherwise stated. Use may be made of any material contained within the thesis for M non-commercial purposes from the copyright holder. Commercial use of material may only be made with the express, prior, written permission of Universiti Putra Malaysia. P Copyright © Universiti Putra Malaysia U T H G I R Y P O C © Abstract of thesis presented to the Senate of Universiti Putra Malaysia in fulfillment of the requirement for the degree of Doctor of Philosophy M MODELLING OF GEOMETRIC CORRECTION AND RELATIVE RADIOMETRIC NORMALIZATION FOR NEAR EQUATORIAL EARTH OBSERVATION SATELLITE IMAGES P By U HAYDER ABD AL-RAZZAQ ABD October 2015 T Chairman : Professor Shattri Bin Mansor, PhD Faculty : Engineering H The near-equatorial earth observation satellite is a new-generation optical satellite that G is expected to cater to the needs of many countries located in or near the equator. It does not move in a fixed row and path during capturing time. Therefore, in each visit, the images exhibit differences depending on the viewing angle, time, illumination, sun zenith and azimuth angles, sensor zenith aInd azimuth angles, altitude, and attitude (roll, yaw, and pitch) of the satellite. These factors result in near-equatorial imagery having: R (a) mis-registration bands, (b) no geometric matching between features located in the sequences of two images even though they are captured from the same strip, and (c) no reflectance (radiometric) matching between features located in two images captured in Y the same area at different times. Conventional modeling does not able to process the near equatorial satellite images. The research aims is study the near equatorial satellite images and develop models that can overcome the difficulties that facing processing of P near equatorial satellite imagery as follow: O (1) To develop band to band registration model for near equatorial image, to overcome the highly nonlinear band shifting of the near equatorial image bands. (2C) To develop geometric correction model for near equatorial images, to reduce the highly geometric correction between the near equatorial imagery. (3) To design and implement remote sensing goniometer to simulate NEqO system images, to simulate the near equatorial satellite image. © (4) To develop relative radiometric normalization model for NEqO images, to normalize the near equatorial images that have been captured at different time. A new technique to overcome band-to-band registration through the automatic generation of control points from satellite images via scale-invariant feature transform (SIFT) is proposed in this study. The SIFT generated control points are utilized to perform registration with first- and second-order polynomials and spline i transformations to correct the misregistration between near-equatorial orbit (NEqO) image bands. The image employed in this study is provided by Malaysian National Space Agency (ANGKASA) it was captured by the RazakSAT satellite. An accuracy M assessment is performed by comparing the result of the proposed method with the result of both automatic and manual transformations using polynomial transformation. The root mean square errors of the first- and second-order polynomial transformations are 4 and 3 m, respectively. Moreover, the spline transformation produces RMSE 1.1 × P 10-6 m. U A new technique is developed to improve the automatic extraction of control points. The technique is then utilized to perform geometric correction of near-equatorial images. The method, which is called refine and, improves scale-invariant feature transform (RI-SIFT). RazakSAT and SPOT-5 images are used. The proposed approach begins by selecting the reference and sensed images. Then, graTyscale and image compression are performed. Automatic control point extraction is then performed to generate control points automatically. The generated control points are refined by using H the sum of absolute difference algorithm (SAD), with the help of an empirical threshold and control point locations to avoid obtaining inaccurate control points. The refined control points are applied in spline transformation to overcome the geometric G errors of near-equatorial orbit (NEqO) images. Validation is then performed by comparing the result of the proposed approach with that of direct geometric correction through the use of polynomial transformation. Results show that the accuracy values obtained from using the refine transformIation (RI-SIFT with spline) and polynomial transformation are 7.08 × 10-9 m and 104 m, respectively. The proposed model exhibits R accuracy and precision. Y A relative radiometric normalization model is implemented in three stages. First, a goniometer is designed and built to simulate near-equatorial images and perform relative radiometric normalization. Second, the goniometer is used to collect images P with different illumination to conduct radiometric normalization. In the third stage, the refine and improve scale-invariant feature transform method is developed to extract automatically radiometric control points over the image bands. The method aims to O acquire the intensity values of the control points to use as pseudo invariant features (PIFs) between the reference and sensed image bands. The next step is to perform statistical linear regression on the control points of the reference and sensed image baCnds to generate regression transformation functions for use in normalizing the sensed image bands to reference image bands. The proposed model is validated by determining the correlation between the normalized and reference image bands. The correlation range is 0.69–0.85 over the bands of the slave image. © ii Abstrak tesis yang dikemukakan kepada senat Universiti Putra Malaysia sebagai memenuhi keperluan untuk ijazah Doktor Falsafah M PEMODELAN PEMBETULAN GEOMETRIK DAN PEMULIHAN RADIOMETRIK RELATIF UNTUK GAMBAR SATELIT PEMERHATIAN BERDEKATAN KHATULISTIWA BUMI P Oleh U HAYDER ABD AL-RAZZAQ ABD Oktober 2015 T Pengerusi : Professor Shattri Mansor, PhD Fakulti : Kejuruteraan H Pemantauan satelit berdekatan khatulistiwa bumi merupakan satelit optikal generasi G baru yang dijangka akan memenuhi keperluan negara-negara yang terletak berdekatan atau di kawasan khatulistiwa. Satelit ini tidak bergerak dalam baris dan laluan yang tetap semasa waktu cerapan. Oleh itu, dalam setiap lawatan, imej-imej yang dipamerkan adalah berbeza bergantung kIepada sudut penglihatan, masa, pencahayaan, nadir matahari dan sudut azimuth, nadir sensor dan sudut azimuth, ketinggian dan ciri- R ciri (roll, yaw dan pitch) satelit tersebut. Faktor-faktor ini menyebabkan imej berhampiran khatulistiwa mempunyai: (a) pendaftaran jalur yang salah, (b) tiada geometri yang sepadan antara ciri-ciri yang terletak diantara urutan kedua-dua imej Y walaupun ianya diambil pada jalur yang sama, dan (c) tiada pantulan (radiometrik) yang sepadan antara ciri-ciri yang terletak di kawasan yang sama pada waktu yang berbeza. Model konvensional tidak mampu untuk memproses imej satelit berdekatan P khatulistiwa ini. Kajian ini bertujuan untuk mengkaji imej satelit berdekatan khatulistiwa dan membangunkan model yang dapat mengatasi kesukaran yang dihadapai semasa memproses imej satelit berhampiran khatulistiwa seperti yang O berikut: (1) Untuk membangunkan model pendaftaran jalur ke jalur untuk imej C berhampiran khatulistiwa, untuk mengatasi peralihan jalur yang sangat tidak linear pada jalur imej berhampiran khatulistiwa. (2) Untuk membangunkan model pembetulan geometri untuk imej berhampiran khatulistiwa, untuk mengurangkan pembetulan geometri antara imej © berhampiran khatulistiwa. (3) Untuk merekabentuk dan melaksanakan goniometer penderia jarak jauh untuk mensimulasi sistem imej NEqO , untuk mensimulasi imej satelit berhampiran khatulistiwa. (4) Untuk membangunkan model radiometrik relatif normal untuk imej NEqO, untuk menormalkn kembali imej berhampiran khatulistiwa yang diambil pada masa yang berbeza. iii Kajian ini mencadangkan satu teknik baru bagi mengatasi pendaftaran jalur ke jalur menerusi penjanaan automatiokautomatik titik kawalan dari imej satelit melalui ciri perubahan skala-tak berubah (SIFT). Titik kawalan yang dijana SIFT digunakan untuk M melaksanakan pendaftaran dengan polynomial tertib pertama dan kedua dan perubahan Spline untuk membetulkan kesalahan pendaftaran antara jalur imej berdekatan orbit khatulistiwa. Imej yang digunakan dalam kajian ini adalah imej satelit RazakSAT yang diperoleh daripada Agensi Angkasa Negara Malaysia (ANGKASA). Penilaian P ketepatan dilakukan dengan membandingkan hasil daripada kaedah yang dicadangkan dengan keputusan yang diperoleh daripada perubahan manual dan automatic menggunakan transformasi polynomial. Punca min ralat kuasa dua pertama dan kedua U untuk transformasi polynomial masing-masing adalah 4 dan 3 m. Selain itu, transformasi spline menghasilkan RMSE 1.1 × 10-6 m. Teknik baru telah dibangunkan untuk meningkatkan pengeluatan tiTtik kawalan secara automatik. Teknik ini kemudiannya digunakan untuk melakukan pembetulan geometri bagi imej berhampiran khatulistiwa. Kaedah ini yang dipanggil penghalusan dan H peningkatan ciri perubahan skala-tak berubah (RI-SIFT). Imej Razaksat dan SPOT-5 telah digunakan. Pendekatan yang dicadangkan bermula dengan memilih rujukan dan imej yang dideria. Kemudian, skala kelabu dan pemampatan imej dilaksanakan. G Pengekstrakan titik kawalan secara automatik kemudiannya dilakukan untuk menjana titik kawalan automatik. Titik kawalan yang dihasilkan ditapis dengan menggunakan sejumlah algoritma perbezaan mutlak (SAD), dengan bantuan ambang empirikal dan titik kawalan lokasi untuk mengelakkan Idaripada mendapat titik kawalan yang tidak tepat. Titik kawalan yang diperhalusi digunakan dalam transformasi Spline untuk R mengatasi kesilapan geometri imej berhampiran orbit khatulistiwa. Pengesahan kemudiannya dilakukan dengan membandingkan hasil daripada pendekatan yang dicadangkan dengan pembetulan geometri secara langsung melalui penggunaan Y transformasi polynomial. Keputusan menunjukkan bahawa nilai-nilai ketepatan diperolehi daripada menggunakan transformasi yang telah diperhalusi itu (RI-SIFT) dan transformasi polynomial adalah 7.08 × 10-9 m dan 104 m masing-masing. Model P yang dicadangkan menunjukkan ketepatan dan kejituan. O Model normalisasi relatif radiometrik ini dilaksanakan dalam tiga peringkat. Pertama, goniometer yang direka dan dibina untuk mensimulasi imej hampir khatulistiwa dan melaksanakan relatif radiometrik pemulihan.Kedua, goniometer yang digunakan untuk mCengumpul imej dengan pencahayaan yang berbeza untuk menjalankan pemulihan radiometrik. Pada peringkat ketiga, penghalusan dan peningkatan ciri skala-tak berubah dibangunkan untuk mengekstrak titik kawalan radiometrik secara automatik antara jalur imej. Kaedah ini bertujuan untuk memperoleh nilai-nilai keamatan titik kawalan untuk digunakan sebagai pseudo ciri tak berubah (PIFs) antara rujukan dan © jalur imej yang dikesan. Langkah seterusnya adalah untuk melaksanakan regresi linear statistik mengenai titik kawalan rujukan dan mengesan jalur gambar untuk menjana fungsi transformasi regresi untuk digunakan dalam menormalkan jalur imej yang dikesan untuk rujukan jalur imej. Model yang dicadangkan disahkan dengan menentukan korelasi antara jalur imej normal dan rujukan. Rangkaian korelasi adalah 0.69-0,85 pada imej slave. iv ACKNOWLEDGEMENT M In the name of Allah, the Most Merciful, the Most Beneficent All praises and thanks are due to my Almighty Allah (SWT) for providing me this opportunity and granting me the capability to proceed successfully. P I would like to express my sincere thanks to Professor Dr. Shattri Bin Mansor, chairman of my supervisory committee, for his dedicated efforts, support, valuable U advice and intellectual guidance during the accomplishment of this research work. I would also like to thank my supervisory committee members, Professor Dr. Noordin Bin Ahmad, Associate Professor Dr. Biswajeet Paradhan for their guidance, assistance, encouragements and constructive comments throughout the period of my study. I also would like to thank Associate Professor Dr. Abdul Rashid Bin MoThamed Shariff and Associate Prof. Dr. Helmi Zulhaidi Mohd Shafri. All staff members of the Geospatial Information Science Research Center and Civil Engineering Department of University H Putra Malaysia. My deepest gratitude goes to my parents, for their help and continuous moral support throughout my study. The help and encouragement provided by my brother Ali, my sisters, my wife and my children (Mohammed, Ali, Fatema and G Rokaia) are greatly appreciated, I R Y P O C © v M P U T H G I R Y P O C © This thesis was submitted to the Senate of Universiti Putra Malaysia and has been accepted as fulfilment of the requirement for the degree of Doctor of Philosophy .The members of the Supervisory Committee were as follows: M Shattri Mansor,PhD P Professor Faculty of Engineering Universiti Putra Malaysia U (Chairman) Noordin Ahmad, PhD Professor Faculty of Engineering T Universiti Putra Malaysia (Member) H Biswajeet Pradhan , PhD Associate Professor Faculty of Engineering G Universiti Putra Malaysia (Member) I R Y P O C BUJANG BIN KIM HUAT, PhD Professor and Dean School of Graduate Studies Universiti Putra Malaysia © Date: vii