高文忠Wen-chung Kao陳聯暘Lien-Yang Chen2019-09-032009-8-82019-09-032007http://etds.lib.ntnu.edu.tw/cgi-bin/gs32/gsweb.cgi?o=dstdcdr&s=id=%22GN0693750080%22.&%22.id.&http://rportal.lib.ntnu.edu.tw:80/handle/20.500.12235/95757色調重現是藉由調整或映射高動態範圍影像，保存原始所看到影像的亮度以及更好的對比度，以獲得真實影像的色調呈現。本論文提出一個影像色調調整的方法，並且整合到我們所開發的穩健的彩色影像處理流程中。藉由影像的巨集邊緣資訊，將巨集邊緣統計分佈圖等化重新分佈，等化分佈的過程中尋求每張圖的色調曲線特性。經由實驗證明，我們提出的色調重現演算法能適應各種場景及環境光源，改善影像的對比及色彩的鮮明度。 除了在PC模擬之外，我們也將演算法移植到嵌入式系統發展版上。針對系統板的特性及加速執行運算速度，改變變數的宣告型態成為整數以及利用查表法技巧。實驗結果證明提出的演算法能適應於一般的嵌入式系統中，並且執行消耗的功率低於個人電腦執行的消耗功率。Tone reproduction attempts to scale or map high dynamic range image data such that the resulting image has preserved the visual brightness and better contrast impression of the original scenes. In this paper, we propose a systematic approach to tone reproduction by histogram equalization of macro edges. The proposed approach has been integrated with a robust image processing pipeline, and the experimental results show that the proposed algorithm is very stable for accommodating diverse illuminants and scenes. In addition to running simulation on a computer, we also have ported the algorithm to an embedded system development kit. Several modifications are implemented such as changing the variables to integer type, and speeding up some time consuming operations by using lookup tables (LUTs.) The experimental results show that the proposed algorithm can be realized on a typical embedded system whose computation power is typically lower than a personal computer.色調重現動態範圍壓縮數位相機直方圖等化Tone reproductiondynamic range reductionhistogram equalizationdigital camera