摘要： 生物在体荧光成像是新兴分子影像技术中性能高、费用低、前景好的一种成像模态。针对生物在体荧光图像的特点和应用需求，提出一种全新的自适应图像分割算法。通过对荧光图像的归一化处理、连通性操作、感兴趣区域区分实现自适应分割。实验结果表明，该算法能够在弱信号、低信噪比、多光源的情况下得到较理想的分割结果，是一种有效的荧光图像分割算法。

关键词: 生物在体荧光成像, 分子影像, 图像分割, 归一化, 自适应分割算法

Abstract: Bioluminescence imaging is regarded as an imaging modality with a high performance, low cost and good prospect in molecular imaging technique. This paper proposes a new adaptive segmentation algorithm, which is based on the characteristics and the application requirements of the bioluminescent images. The adaptive segmentation is realized by performing the normalized processing, connectivity operation and the interested regions distinguishing on bioluminescent images. Experimental results show that this algorithm can get better segmentation results in the case of weak signal, low signal-to-noise ratio and multiple light sources, so it is a kind of effective segmentation algorithm of bioluminescent images.

Key words: bioluminescent image, molecular image, image segmentation, normalization, adaptive segmentation algorithm

中图分类号: 

• TP301.6