论文部分内容阅读
提出了一种将背景区域感知与目标特征分析相结合的自动提取图像目标的层次化数据处理流程;探讨了线性预测估计和非线性均值滤波两种常用背景估计技术和自适应空间滤波的形态学背景感知方法。实验测评表明,形态滤波算法对背景的自适应感知能力强,能真实贴切地反映图像背景的起伏变化规律。
This paper presents a hierarchical data processing flow of automatic extraction of image objects by combining background area perception with target feature analysis. Two commonly used background estimation techniques, such as linear prediction estimation and nonlinear mean filtering, and morphology of adaptive spatial filtering Background perception method. Experimental results show that the morphological filtering algorithm has strong adaptive perception to the background and can reflect the changing regularity of the background of the image in a real and appropriate way.