重庆理工大学学报（自然科学版）
重慶理工大學學報（自然科學版）
중경리공대학학보（자연과학판）
JOURNAL OF CHONGQING INSTITUTE OF TECHNOLOGY
2015年
5期
71-77
,共7页
遥感图像%分类%视觉词袋模型%面向对象
遙感圖像%分類%視覺詞袋模型%麵嚮對象
요감도상%분류%시각사대모형%면향대상
remote sensing image%classification%bag-of-visual-words model%object-based
研究了基于视觉词袋模型的单幅全色遥感图像的分类方法。首先提取图像边缘特征与图像区域，以旋转不变纹理基元作为视觉词汇，再采用面向对象方法进行分类。仿真结果表明：该方法具有较高的分类精度，且具有用时成本较低和适应性强的优点。
研究瞭基于視覺詞袋模型的單幅全色遙感圖像的分類方法。首先提取圖像邊緣特徵與圖像區域，以鏇轉不變紋理基元作為視覺詞彙，再採用麵嚮對象方法進行分類。倣真結果錶明：該方法具有較高的分類精度，且具有用時成本較低和適應性彊的優點。
연구료기우시각사대모형적단폭전색요감도상적분류방법。수선제취도상변연특정여도상구역，이선전불변문리기원작위시각사회，재채용면향대상방법진행분류。방진결과표명：해방법구유교고적분류정도，차구유용시성본교저화괄응성강적우점。
The method of panchromatic remote sensing image classification based on bag-of-visual-words model was studied. Firstly,we extracted the feature of boundaries and the segments. Secondly, we used the rotation invariant texture feature as visual vocabulary. Finally,we applied the method of object-based for classification. The simulation results demonstrate that the proposed method has higher classification accuracy and has the advantages of low time-cost and strong adaptability.
