基于流型学习的地面结霜现象检测  

Automated surface frost detection based on manifold learning

作　　者：朱磊[1,2] 曹治国[1] 肖阳[1] 李肖霞[3] 马舒庆[3] 

机构地区：[1]华中科技大学自动化学院,武汉430074 [2]武汉科技大学信息科学与工程学院,武汉430081 [3]中国气象局大气探测中心,北京100081

出　　处：《计算机应用》2015年第3期854-857,871,共5页journal of Computer Applications

基　　金：国家公益性行业(气象)科研专项基金资助项目(GYHY200906032)

摘　　要：针对日常地面气象观测中近地面结霜现象仍需要依靠人工观测来完成的问题,提出了一种基于计算机视觉的结霜现象自动化观测方法。在实时检测中,首先,结合人工标记获取的离线结霜图像样本和实时获取的图像样本构造k近邻图模型;其次,以结霜图像样本为查询节点并通过流型学习方法在图模型上对实时图像样本进行排序,进而获取候选结霜区域;最后,根据结霜和非结霜图像样本在线训练支持向量机(SVM)分类器并对候选结霜区域进行二次判定。在标准化气象观测站实施的实验结果显示,对比同期人工观测记录,该算法对结霜现象的检测正确率达到了87%,具有潜在的业务化前景。As an important component of the surface meteorological observation, the daily observation of surface frost still relies on manual labor. Therefore, a new method for detecting frost based on computer vision was proposed. First, a k-nearest neighbor graph model was constructed by incorporating the manually labeled frosty image samples and the test samples which were acquired during the real-time detection. Second, the candidate frosty regions were extracted by rating those test samples using a graph-based manifold learning procedure which took the aforementioned frosty samples as the query nodes. Finally, those candidate frosty regions were identified by an on-line trained classifier based on Support Vector Machine （ SVM）. Some experiments were conducted in a standardized weather station and the manual observation was taken as the baseline. The experimental results demonstrate that the proposed method achieves an accuracy of 87% in frost detection and has a potential applicability in the operational surface observation.

关 键 词：地面气象观测 结霜现象 计算机视觉 流型学习 在线训练 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]