摘要
本文针对建筑单体化三维模型贴图中存在的亮度不均匀和噪点问题,提出了一种基于多尺度视网膜增强色彩恢复(MSRCR)和非局部均值(NLM)去噪算法的自适应优化方法。为消除大范围黑色区域对后续处理的影响,研究中创建了掩码文件,并计算贴图法线,筛选出法线水平的贴图,过滤掉亮度适中的顶部和底部贴图。在特征提取阶段,本文计算了每组水平贴图的平均亮度和亮度直方图的阈值,以便更精确地进行亮度增强。采用基于MSRCR的亮度增强算法,对筛选出的低亮度贴图进行有效提亮。同时,利用NLM算法对贴图进行去噪处理,确保在去除噪点的过程中,原始细节和色彩得到保留。通过对处理结果进行主观和客观的评价,结果表明,本文方法显著提高了模型贴图的亮度和清晰度。在实际应用中,本文方法展示了其在提升实景三维模型贴图质量方面的高实用价值,为建筑三维建模领域提供了新的解决方案。
This paper addresses issues of uneven brightness and noise in the texture mapping of monolithic three-dimensional(3D)building models.An adaptive optimization method is proposed,based on the multi-scale retinex with color restoration(MSRCR)and non-local mean(NLM)denoising algorithms.To eliminate the influence of large black areas on subsequent processing,a mask file was created,and texture normals were calculated to filter out horizontal textures.Textures with mod⁃erate brightness at the top and bottom were excluded.In the feature extraction phase,the average brightness and brightness histogram thresholds for each set of horizontal textures were calculated to achieve more precise brightness enhancement.A brightness enhancement algorithm based on MSRCR was applied to effectively brighten the low-brightness textures selected for improvement.At the same time,the NLM algorithm was used to denoise the textures,ensuring that the original details and colors were preserved during the noise removal process.Both subjective and objective evaluations of the results show that the proposed method significantly improves the brightness and clarity of the model textures.In practical applications,the method demonstrates high practical value in enhancing the quality of texture mapping for real-world 3D models,offering a new solution for the field of architectural 3D modeling.
作者
曹明亮
赵凌美
王淼
CAO Mingliang;ZHAO Lingmei;WANG Miao(Beijing Institute of Surveying and Mapping,Beijing,100038,China;School of Land Science and Technology,China University of Geosciences Beijing,Beijing 100083,China)
出处
《北京测绘》
2025年第9期1336-1343,共8页
Beijing Surveying and Mapping
基金
2024年北京优秀青年工程师创新工作室项目(2024-13)
北京高等学校卓越青年科学家项目(JJWZYJH-01201910003010)。
