像素级融合方法中常出现色彩突变或色彩失真现象。通过分类信息对融合进行约束可以部分消除目标地物边界的这些现象。然而传统的基于像素分类的影像融合方法由于分类中的“椒盐效应”,导致融合效果受到一定的影响和限制。采用面向对象分类约束的方法对该融合方法进行改进。首先采用面向对象分类方法进行影像分类,解决了基于像素分类中的“椒盐效应”问题;其次将分类结果作为影像融合的约束条件,利用色度饱和度明度(HSV)变换进行融合;最后将该方法的结果与多种融合方法的结果进行定量比较,发现该方法除在目视上取得很好的增强效果外,在信息熵、方差等指标上也取得了很好的效果。 Pixel fusion methods often appear sudden changes in color or color distortion. Constraints on the fusion by the classification information can partially eliminate these phenomena at the boundary of the object. However, the traditional pixel-based image fusion method has some limitations and effects on the fusion effect due to the “salt and pepper effect” in classification. The method of object-oriented classification is used to improve the fusion method. Firstly, object-oriented classification method was used to classify the images, which solved the problem of “salt and pepper effect” based on pixel classification. Secondly, the classification results were used as the constraint of image fusion and fused by HSV (Color Saturation Lightness). Finally, The results of the method are quantitatively compared with those of many fusion methods. It is found that this method not only achieves good enhancement effect visually, but also achieves good results on the indexes of information entropy and variance.