CAJ | 학술논문

为了实时高效地对车载红外视频图像进行彩色化处理,提出了一种在Lab颜色空间基于调色板的分类彩色化方法.对增强和融合处理后的红外图像采用K均值聚类方法得到聚类中心,再利用Fisher评价函数获得分类界限以对红外视频图像进行图块景物分割,在Lab颜色空间中采用分类彩色化方法对各类别景物赋予适当的色彩.试验结果表明,采用调色板的分类彩色化方法具有更高的效率,得到的图像色彩符合人眼的视觉特性且便于夜间红外图像的彩色化实现.
위료실시고효지대차재홍외시빈도상진행채색화처리,제출료일충재Lab안색공간기우조색판적분류채색화방법.대증강화융합처리후적홍외도상채용K균치취류방법득도취류중심,재이용Fisher평개함수획득분류계한이대홍외시빈도상진행도괴경물분할,재Lab안색공간중채용분류채색화방법대각유별경물부여괄당적색채.시험결과표명,채용조색판적분류채색화방법구유경고적효솔,득도적도상색채부합인안적시각특성차편우야간홍외도상적채색화실현.
This paper puts forward a color classification method based on color palette technology in the Lab color space so as to carry out efficiently the color processing of vehicle-mounted infrared video images in time. The system firstly applied k-means clustering method to the preprocessed infrared images and obtained the clus-tering center. Then it segmented the block scenery of infrared video images with classification boundaries ob-tained by Fisher evaluation function. The appropriate color was thus given to different scenery frames by the color classification method. Experiments show that the proposed color classification method has much higher efficiency which can obtain realistic image color fit for human vision and is convenient for achieving coloriza-tion of infrared video images at night.