以攝影測量方式建立無人機影像曜光模式之研究

Abstract

當光線出射物質表面之角度和衛星或飛行器的感測器達成特定幾何關係，且光線入射表面之粗糙程度與波長達到特定關係時，便會在遙測影像中出現太陽曜光(Sun Glint / Sun Glitter)現象。太陽曜光會對遙測數據造成程度不等之影響，故過去衛星影像領域中已發展出多種太陽曜光偵測與改正的方法。但過去應用於衛星及航照影像的預估或改正模式，是否能有效應用於空間解析度高的無人機影像，仍需進一步驗證。儘管無人機影像領域中已開始有與改正曜光相關的研究正在進行，但目前應用與分析的軟體上還未有太多曜光改正的支援，且多數改正模式需搭配可見光以外之波段進行。本研究欲於無人機於航線規劃階段，透過研究建立之曜光預估模式，瞭解曜光可能的出現情形，以減少或避免飛行任務結束後出現在航照影像中的曜光影響情形。對研究區域建立地表、太陽及攝影站之空間關係後，分別以過去研究所整理出，以及本研究嘗試簡化自前者概念的計算方式，進行曜光預估的計算。並進一步設計實驗，將實驗成果分別以影像、統計及實體空間的角度，呈現目標飛行時段的曜光結果，以得知目標飛行時段下出現最少曜光的時段，降低後續去除曜光作業時間，並為後續無人機影像曜光處理之相關研究參考。根據研究之模擬，使用攝影測量方式建立之曜光預估模式可使使用者於規劃階段得知曜光於整體影像蒐集過程之分佈，並可依循時間、外方位元素調整的要點，使一日之中可飛行之時段，帶來較高有效影像蒐集效率之飛行規劃。Sun glint occurs when the angle at which sunlight hits and reflects off a surface aligns perfectly, creating a mirror-like reflection. This intense glare can obstruct crucial details in an image, making it challenging to distinguish objects or features on the Earth's surface. Nowadays, Unmanned Aerial System (UAS) imagery products also suffer from blurring and degradation caused by sun glint, resulting in reduced image contrast, color fidelity in photogrammetry, and compromised radiometric values in remote sensing applications. Various techniques, including detection methods and specialized algorithms, are used to minimize sun glint's impact in satellite imagery. However, it remains uncertain if the processing methods in satellite imagery can effectivelyapplied to images with higher spatial resolution.By establishing the Geometric Relationship between the earth surface, sun, and sensor, a threshold will be set to determine the presence or absence of sun glint. The threshold determination will be based on past research findings and also on thresholds developed within this study, providing distinct threshold values for different computation methods. The research aims to grasp the impact of sun glint during various time periods by calculating the results before conducting the flight, with the objective of identifying the time period with the least occurrence of sun glint. Utilizing photogrammetric techniques to establish a sun glint prediction model provides users with insights into the distribution of sun glint throughout the entire image acquisition process during the planning phase. This enables the feasibility of planning flight schedules during periods of the day that offer higher efficiency in capturing useful images through adjusting the timing and exterior orientation.