تجزیه وتحلیل و مقایسۀ روش های مرسوم زمین مرجع کردن عکس های هوایی (مقاله علمی وزارت علوم)
درجه علمی: نشریه علمی (وزارت علوم)
آرشیو
چکیده
عکس های هوایی برای سالیان متمادی، پایه و اساس بسیاری از پژوهش های کاربردی انجام شده بر روی منابع سرزمین، ازجمله مطالعات خاک شناسی بوده اند. هم زمان با پیشرفت های فناوری و ورود سامانه های اطلاعات جغرافیایی به مطالعات سرزمین به منظور تجزیه وتحلیل های جغرافیایی، زمین مرجع کردن عکس هوایی در این سامانه ها موردتوجه قرار گرفت. درست و دقیق انجام شدن این فرایند، به دلیل ماهیت عکس های هوایی، اعوجاج و جابه جایی پدیده ها نسبت به موقعیت واقعی آنها بسیار ضروری و مهم است. بر این اساس، هدف از این پژوهش، دستیابی به روشی مناسب برای زمین مرجع کردن عکس های هوایی است که افزون بر ساده و کم هزینه بودن، با دقت باشد. برای این منظور، روش های مرسوم زمین مرجع کردن عکس های هوایی (روش های تبدیل درجه اول تا درجه سوم، اسپیلاین، تصویری و اصلاح قائم سازی) در دو منطقه متفاوت ازنظر ناهمواری ها (منطقه ی هموار شهرکرد و منطقه ی تپه ماهوری چالشتر)، در نرم افزار ILWIS3.3 وArcGIS10.7 استفاده و خطای روش های مختلف از طریق محاسبه شاخص های کمینه و بیشینه، میانگین (ME)، انحراف معیار و جذر مربع خطاها (RMSE) تجزیه وتحلیل شد. نتایج حاکی از آن بود که برای منطقه هموار، خطای روش های مختلف، تغییرات کمتری نسبت به منطقه ناهموار داشت؛ به گونه ای که با افزایش ناهمواری ها، مقدار خطای روش ها به شدت افزایش یافت و نشان دهنده تفاوت چشمگیری بود. براساس نتایج به دست آمده، روش اصلاح قائم سازی صرف نظر از زمان و داده های موردنیاز، دقیق ترین روش زمین مرجع کردن بود (35 = ME و 38 = RMSE)؛ اما با در نظر گرفتن زمان و داده هایی که برای این روش لازم بود، روش اسپیلاین مناسب ترین روش تشخیص داده شد. در روش اخیر نه تنها دقت مقبولی وجود دارد، مشکلات روش اصلاح قائم سازی را ازجمله نیاز به داده های حاشیه ای عکس هوایی و لایه مدل رقومی ارتفاع نیز ندارد.Analysis and Comparison of the Conventional Methods of Georeferencing of Aerial Photos
Aerial photos are the basis of many researches, both applied and executive works, related to the earth. Applications of aerial photos, especially for GIS-based analysis, aerial photo georeferencing, and/or geometric correction are essential. Various methods have been proposed for georeferencing. The aim of this study was to achieve a suitable simple method with a low cost and acceptable accuracy. For this purpose, conventional georeferencing methods, including first- to third-order polynominal transformtions, spline transformtion, projective transformtion, and orthorectification transformtion were used in ILWIS3.3 and ArcGIS10.7 for two areas (flat area of Shahrekord and rugged area of Chaleshtor). Errors of the different methods were analyzed based on minimum and maximum errors, Mean of Error (ME), standard division, and Root Mean Square Error (RMSE). The results showed that the errors of the varied methods had fewer changes for the flat compared to the rugged area. Yet, with the increasing relief of land surface, the errors increased sharply and showed a significant difference. Based on the results, the most accurate georeferencing method was orthorectification method (ME=35 and RMSE=38). However, considering the time and data required for the orthorectification method, the most suitable georeferencing method was identified as the spline method, which had acceptable accuracy, but not the problems of the orthorectification method, such as marginal information of the aerial photos and digital elevation models. Introduction: The potential of aerial photographs as a tool in the research of land resources has long been recognized. It has been the basis of many applied types of research, including soil survey studies. Despite the informative potential of aerial photographs, so far, their usage has been scarce due to the processing difficulty related to the lack of some key information. Fortunately, a recently introduced technology, such as Geographic Information System (GIS), has shown to be able to overcome the classic photogrammetric limitations. Generally, , these photos must be converted from analog to digital forms to be used as a basis for mapping of soil resources. Thus, they have to georeferenced and then imported into GIS. However, the accuracy of this process is very important due to the nature of aerial photographs and distortion caused by moving the phenomena to their actual positions. Therefore, the research design achieved an acceptable method of georeferencing with high accuracy and low cost, besides not being time-consuming. Materials and Methods: For this purpose, two study areas were selected from the roughness point of view (a flat area and a hilly area). Then, 6 transformation methods were applied for georeferencing of aerial photographs, including first-, second-, and third-order polynomial transformations, projective transformation, spline transformation, and orthorectification. Afterwards, 15 ground control points and 15 specific points on the aerial photographs were selected for measuring the error values. They included the points that were clearly and accurately recognizable on both the aerial photographs and Google Earth images, such as roads, buildings, waterways, peaks, etc. The coordinates of all the points were obtained based on the Universal Transverse Mercator (UTM) coordinate system using Google Earth images. The measurement error for each of these points was obtained based on the Euclidean distance between each point on the georeferenced aerial photographs and the Google Earth images. Finally, Descriptive statistics, such as minimum and maximum errors, Mean of Error (ME), standard deviation, and Root Mean Square Error (RMSE) were utilized to compare the errors of the 6 used georeferencing methods and introduce the best one. Georeferencing was done by using ILWIS 3.3 and ArcGIS10.7 and the calculations were performed via Excel. Results and Discussion: The results revealed that the errors of the different methods had fewer changes for the flat compared to the rugged area, but with the increaing relief of land surface, the errors increased sharply and showed a significant difference. Based on the results, orthorectification was the most accurate method of transforming the aerial photographs into the ground coordinate system (MR=35 and RMSE=38). However, this method required many data and was a time-consuming and costly method. Therefore, the most suitable approach to transforming the aerial photographs into the ground coordinate system, especially when there was not enough information and time for orthorectification, could be the spline method, which had acceptable accuracy for both flat and uneven areas. Keywords: geometric correction, distortion, displacement, Geographical Information System (GIS) References: - Bannari A., Karl S., Catherine Ch. and Shahid Kh. (2015). Spatial Variability Mapping of Crop Residue Using Hyperion (EO-1) Hyperspectral Data. Remote Sensing , 7, 8107-8127. - Casson, B., Delacourt, C., Baratoux, D., & Allemand, P. (2003). Seventeen years of the “La Clapiere” landslide evolution analysed from ortho-rectified aerial photographs. Engineering Geology , 68, 123–139. - Chen, J. J., Huang, C. L., Wu, Y. G., & Wu, P. H. (2021). Influence Factors of 3D Modeling with Aerial Images. International Journal on Computer, Consumer and Control, 10 (2), 1-11. - Chintan . D ., Rahul . J., Srivastava. S., (2015). A Survey on geometric correction of satellite imagery. International journal of computer applications , 116 (12). - Hackeloeer, A., Klasing, K., Krisp, J. M., & Meng, L. (2014). Georeferencing: a review of methods and applications. Annals of GIS , 20(1), 61-69. - Karsli, F., & Dihkan, M. (2010). Determination of geometric deformations in image registration using geometric and radiometric measurements. Scientific Research and Essays 5(3), 260-274. - Lee, H.L., Beng, Y.L., Yin, C.W., & Wai, S.C. (2013). Aerial Images Rectification Using Non-parametric Approach. Journal of Convergence 4(2), 15-21. - Liu D, Zhou G, Huang J, Zhang R, Shu L, Zhou X, Xin CS. (2019). On-Board Georeferencing Using FPGA-Based Optimized Second-Order Polynomial Equation. Remote Sensing , 11(2):124. https://doi.org/10.3390/rs11020124 - McNish, I. G., & Smith, K. P. (2022). Oat crown rust disease severity estimated at many time points using multispectral aerial photos. Phytopathology , 112(3), 682-690. - Pha S.H. and Takeuchi W. (2017). Effect of GCPs in distribution and location on geometric correction of corona satellite image. Coordinate Magzine , XIII(07), 14-19. - Powers, P. S., Chiarle, M., & Savage, W. Z. (1996). A digital photogrammetric method for measuring horizontal surficial movements on the Slumgullion earthflow, Hinsdale County, Colorado. Computers and Geosciences , 22, 651–663. - Richards, J. A., & Jia, X. (1999). Remote sensing digital image analysis. An introduction . Berlin: Springer. - Riquelme A., Soldato M. D., Tomás R., Cano M., Bordehore L. J. and Moretti S. 2019. Digital landform reconstruction using old and recent open access digital aerial photos. Geomorphology , 329, 206-223. - Rocchini D., A. Di Rita, 2005, Relief effects on aerial photos geometric correction. Applied Geography, 25, 159–168. - Rossiter, D. G., and Tomislav Hengl. (2004). Technical note: Creating geometrically-correct photo-interpretations, photomosaics, and base maps for a project GIS . Enschede, NL: ITC, Department of Earth System Analysis. - Russ, J. C. (2002). The image processing handbook (3rd ed). Boca Raton, FL: CRC Press. - Sameen M. I., Pradhan B., Aziz O. S. (2018). "Classification of Very High-Resolution Aerial Photos Using Spectral-Spatial Convolutional Neural Networks", Journal of Sensors , 2018, 1-12. - Sempio, J. N. H., Aranas, R. K. D., Lim, B. P., Magallon, B. J., Tupas, M. E. A., and Ventura, I. A. (2019). Assessment of different image transformation methods on diwata-1 smi images using structural similarity measure, The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. , XLII-4/W19, 393–400, https://doi.org/10.5194/isprs-archives-XLII-4-W19-393-2019. - Tao, C.V.; Hu, Y. (2001). Use of the Rational Function Model for Image Rectification. Canadian Journal of Remote Sensing , 27(6), 593–602. doi:10.1080/07038992.2001.10854900. - Toutin T. (2004). Review article: Geometric processing of remote sensing images: models, algorithms and methods, International Journal of Remote Sensing , 25(10), 1893-1924 - Wang M., J. Hu, M. zhou, J.M. Li, Z. Zhang. (2013). Geometric Correction Of Airborne Linear Array Image Based On Bias Matrix. International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences , Hannover Workshop, 21 – 24 May, Hannover, Germany. Figures and Tables - Fig. 1: Study areas and locations of the aerial photos of Shahrekord and Chaleshtar - Fig. 2: A 3-dimensional (south to north) view of the aerial photos of the relief areas Shahrekord (top) and Chaleshtar (bottom) - Fig. 3: The original aerial photograph of Shahrekord Area with ground control points (triangles) and error measurement points (circles) - Fig. 4: The original aerial photograph of Chaleshtar Area with ground control points (triangles) and error measurement points (circles) - Table 1: Error calculation by affine (first-order) transformation method for the aerial photos of Shahrekord - Table 2: Measured error statistics for the aerial photos of Shahrekord - Table 3: Measured error statistics for the aerial photos of Chaleshtor - Fig. 5: The effects of the different georeferencing methods on the aerial photos of Shahrekord (Transformations: 1) affine (first order); 2) second order; 3) third order, 4) projective; 5) spline; and 6) orthorectification with first order)
تبلیغات
