周高動態範圍：高動態範圍影像暨色彩平衡工具

Abstract

隨著智慧型手機攝影技術的快速發展，用戶對於照片品質的要求日益提高。特別是在開發新型手機相機模組的過程中，如何確保由開發中手機所拍攝的照片與參考機拍出的照片品質一致，成為一項重要的挑戰。本研究旨在強化測試流程，通過改進影像處理技術，使得測試手機所拍攝的影像能更接近參考機的表現。研究涵蓋了高動態範圍（HDR: High Dynamic Range）影像處理、自動曝光（Auto Exposure, AE）調整，以及自動色調映射（Color Automatic Tone Mapping）等核心技術。我們提出了一套以影像特徵為基礎的自動化測試流程，包含色彩校正、對比度優化與曝光平衡的調整策略，並透過一系列實驗驗證了所提方法的有效性。結果顯示，該方法能顯著提升測試手機影像的品質，使其在亮度、色彩及細節呈現上接近參考機標準，為未來手機攝影技術的開發與測試提供了可靠的參考框架。 本研究包含兩個主要階段：高動態範圍參數校正、色彩平衡參數校正。首先，我們用自己的手機拍攝照片，與對比的手機拍攝的照片做比較。將兩張照片套用至 ChouHDRI，我們可以獲得高動態範圍的參數校正。若是要將照片與對比機的照片做色彩比較，ChouHDRI可以獲取色彩相關的修正參數。在高通的測試軟體上修改這些參數，再重新拍攝後，就可以獲得兩張曝光與色彩都更加相似的照片。 ChouHDRI為手機開發者及測試者提通一個更快速便利的工具，使得測試開發階段可不再仰賴人眼辨識，並提供更準確的校正參數參考。並且可依據需求，提供不同種對比機的照片模擬。

With the rapid development of smartphone photography technology, users’ demands for photo quality are steadily increasing. In particular, during the development of new smartphone camera modules, ensuring that the photos taken by a developing smartphone match the quality of those captured by a reference camera has become a significant challenge. This study aims to strengthen the testing process by enhancing image processing techniques so that the images captured by the test smartphone more closely resemble the performance of the reference camera. The research covers core technologies such as High Dynamic Range (HDR) image processing, Auto Exposure (AE) adjustment, and Color Automatic Tone Mapping. We propose an image feature-based automated testing process, which includes adjustment strategies for color calibration, contrast optimization, and exposure balance, and we validate the proposed method through a series of experiments. The results indicate that this method can significantly improve the quality of images captured by the test smartphone—achieving brightness, color, and detail presentation similar to the reference standard—and provide a reliable framework for the development and testing of future smartphone photography technologies. This study comprises two main stages: HDR parameter calibration and color balance parameter calibration. First, we compare photos taken by our own smartphone with those captured by a reference smartphone. By applying both photos to ChouHDRI, we obtain the HDR parameter calibration. When comparing the color of the photos with that of the reference camera, ChouHDRI can acquire the related color correction parameters. After modifying these parameters in Qualcomm’s testing software and retaking the photo, we can achieve two images that are even more similar in both exposure and color. ChouHDRI provides a faster and more convenient tool for smartphone developers and testers, eliminating the need to rely on human visual inspection during the testing and development stages, and offering more accurate calibration parameters. Additionally, it can simulate images from various reference cameras based on specific requirements.