摘要
针对传统的高动态范围图像合成方法不能适应动态光照的问题,提出了基于相机阵列的不同曝光的多幅图像的配准及高动态范围图像合成方法。首先利用相机阵列获取不同曝光图像,结合相机阵列标定参数,采用光场合成孔径理论对图像进行配准,并对配准后的图像作中值位图进行二次配准。根据拟合出的各相机的光照响应曲线,进而将二次配准后的不同曝光的图像合成为一幅高动态范围图像。实验表明,该方法可以有效地在动态光照下合成高动态范围图像,取得了不错的效果。
For traditional high dynamic range （HDR） image synthesis method could not adapt to the dynamic illumination, the paper proposed a camera array-based different exposure images alignment and HDR image synthesis method. Firstly it cap- tured different exposure images with the camera array whose calibration parameters had been calculated out, according to light field synthetic aperture photography theory, aligned different exposure images, and then did a second alignment to aligned ima- ges with the median threshold bitmap （MTB）. According to different cameras＇ response curves fitted, it generated a HDR ima- ge with different exposure images second aligned. The experiments show that the proposed method can effectively synthesize HDR images in the dynamic illumination conditions, and achieves good results.
出处
《计算机应用研究》
CSCD
北大核心
2013年第9期2858-2860,2864,共4页
Application Research of Computers
基金
国家"863"计划资助项目(2012AA011803)
