基于视觉显著性模型的地域建筑肌理特征识别  

Research on Regional Architectural Texture Feature Recognition Based on Visual Salience Model

作　　者：梁楠 LIANG Nan(Anhui Polytechnic University, Wuhu 241000, China)

机构地区：[1]安徽工程大学建筑工程学院,安徽芜湖241000

出　　处：《洛阳理工学院学报（自然科学版）》2021年第4期27-33,46,共8页Journal of Luoyang Institute of Science and Technology：Natural Science Edition

基　　金：安徽高校人文社会科学重点研究项目基金(SK2020A0232).

摘　　要：在地域建筑肌理识别中运用视觉显著性模型,可有效增强地域建筑肌理识别效果。通过直方图增强技术增强地域建筑肌理图像,采用视觉显著性算法提取地域建筑肌理特征的视觉显著性,结合阈值分割的最大类间差法分割视觉显著性图像,实现目标的精准识别。最终采用BoVW(Bag of Visual Words)模型的特征因子迭代和支持向量机的函数分类,完成地域建筑肌理特征识别。经实验检验,构建的基于视觉显著性模型地域建筑肌理特征识别方法可以有效增强识别图像,边界轮廓分隔效果与视觉显著性效果良好,不同建筑物场景的肌理特征识别率均值达到0.814,具有较高稳定性。The application of visual salience model to the regional architectural texture feature recognition can greatly improve the recognition effect.The histogram enhancement is used to improve the regional architectural texture image and the visual saliency of the texture feature is extracted based on the visual saliency algorithm,and segment the visual saliency image with the maximum difference between classes method of threshold segmentation,to obtain the accurate recognition.Through the feature factor iteration based on the bag of-visual-words(BOVW)model and the function classification of support vector machine,the texture feature recognition of regional architecture is completed.The experiment shows that the model can effectively enhance the recognition image,with clear boundary contour of the segmentation and good visual saliency effect.The average recognition rate of texture features in different buildings reaches 0.814,with the high stability.

关 键 词：视觉显著性 地域建筑 特征识别 特征提取 图像增强 目标分割 

分 类 号：TU238[建筑科学—建筑设计及理论]