基于MRF的航空影像理解的知识获取
The Knowledge Acquisition Based on Markov Random Field for Aeroimage Understanding
-
摘要: 讨论了基于MRF影像理解的3种知识——同类影像的距离最短、近邻像元个别特征参数的显著性和影像的方向性在航空影像理解中如何获取和应用。通过航空影像试验,表明影像理解的正确率大大提高。Abstract: This paper discusses the three kinds of knowledge based on Markov random field:the distance between the same class images is the shortest;some feature parameters of neighbour pixels are evidence;and images have direction feature.It also presents how the three kinds of knowledge can be acquired and applied in image understanding.Experiments show that these knowledge can help to improve the correct ratio of image understanding.
下载:
