在電腦圖形學與電影攝影術中,比起普通數位影像技術,高動態範圍成像是用來補捉更大曝光動態範圍(即更大的明暗落差)的一門技術。高動態範圍成像的目的就是要能正確地表示在真實世界場景中,從太陽光直射到最暗的陰影之間這樣廣的亮度範圍。高動態範圍影像發展可使用不同的方法生成,但其中最著名的一個方法是透過估算輝度。由於所使用單位都是基於□這樣的實際物理單位,而在科技日新月異的現在,相機也發展到了可調整光圈、快門、iso...更多係數的數位相機,透過得到更多不同動態範圍的影像,能夠更加精準的重現真實場景的光線於數位影像顯示設備上。數位相機上的EV值取決於光圈、快門、iso的組合,可配合不同的場景需求來透過不同的設定值來得到相同的輝度,但是無法確定數位相機上的快門時間是否為它顯示般的準確;本研究將透過控制光圈、快門的設定來營造相同的輝度,再經過量測其輝度來推演各個不同快門時間所需要的修正公式(K函式),以補償相機在拍攝上的不足、讓高動態範圍還原技術更加精準。
Radiance information is an important factor in processing high dynamic range image. It was not so easy to recover the true measurement of scene radiance from a sheet of film since the exposure settings may not be available. Nowadays with the stored exposure information in the digital camera, one can obtain a true measurement of relative radiance easier when we photograph a scene with digital cameras. This research proposes a algorithm to recover high dynamic range radiance maps from a series of differently exposed images of the same scene. At first, several training images are used to estimate the response function of the camera in capturing images. Then, this method merges multiple photographs into a single radiance map with the known response function. High dynamic range radiance map whose pixel values are proportional to the true radiance values in the scene can therefore be obtained. This paper demonstrates a key finding in this method.