Imtiaz Mohammad Shamim, Wahid Khan A
Annu Int Conf IEEE Eng Med Biol Soc. 2014;2014:3905-8. doi: 10.1109/EMBC.2014.6944477.
This paper presents an image enhancement and space-variant color reproduction method based on adaptive sigmoid function for endoscopic image. At first, using YCBCR conversion matrix, the color image is separated into luminance and chrominance components. The adaptive sigmoid function with two controlling parameters is applied on the uniformly distributed luminance pixels. The space-variant color reproduction generates new chrominance components by transferring and modifying old chrominance based on texture information. Finally, new luminance and chrominance components are converted into RGB color image. The proposed method highlights some of the tissue and vascular characteristics as well as pit patterns in lesion and polyp. The performance of the proposed scheme is compared with other related methods in terms of image quality, focus value, efficiency of color reproduction and statistic of visual representation.
本文提出了一种基于自适应Sigmoid函数的内窥镜图像增强及空间可变颜色再现方法。首先，利用YCBCR转换矩阵将彩色图像分离为亮度和色度分量。将具有两个控制参数的自适应Sigmoid函数应用于均匀分布的亮度像素。空间可变颜色再现通过基于纹理信息传递和修改旧色度来生成新的色度分量。最后，将新的亮度和色度分量转换为RGB彩色图像。所提方法突出了病变和息肉中的一些组织、血管特征以及凹陷图案。在所提方案的性能方面，从图像质量、聚焦值、颜色再现效率和视觉表示统计等方面与其他相关方法进行了比较。
