光学学报, 2020, 40 (8): 0810001, 网络出版: 2020-04-13   

基于局部异常因子的近地全天时星图小波去噪 下载: 960次

Wavelet Denoising of Near-Earth All-Day Star Map Based on Local Outlier Factor
吴强 1,2,*张锐 1,**
作者单位
1 中国科学院微小卫星创新研究院, 上海 201203
2 中国科学院大学, 北京 100049
摘要
星图信噪比是影响星敏感器拍摄星图中星点提取精度的重要因素。软阈值等去噪方法在处理近地面全天时星图时其阈值选取问题引起的噪声残留会影响星点质心的提取精度。针对这一问题,提出一种加权局部异常因子(LOF)的近地全天时星图小波去噪方法。该方法将局部异常因子算法应用于星图的小波去噪中,实现了不依赖阈值的近地全天时星图去噪。以地面真实拍摄的星图作为原始数据,使用峰值信噪比(PSNR)及局部峰值相对误差(LPVRE)对不同去噪方法处理后的星图去噪效果进行对比分析。实验结果表明,本文方法相较传统均值滤波和小波阈值去噪,提高了峰值信噪比,降低了局部峰值相对误差,能较好地去除背景噪声并较好地保留目标信息。
Abstract
The signal-to-noise ratio (SNR) of a star map is an important factor affecting the accuracy of star point identification. For the threshold denoising methods, the noise residual caused by the threshold selection problem in the ground all-day star map affects the accuracy of the star point centroid extraction. This study proposes a near-earth all-time star map wavelet denoising method based on the local outliers factor. The proposed method applies the local outliers factor algorithm to the wavelet denoising of the star map to perform the denoising of the ground all-time star map without threshold. Herein, the real star map is considered as the original data, and the peak SNR (PSNR) and local peak value relative error (LPVRE) are used to compare and analyze the denoising effect of the star map processed using different denoising methods. Results show that compared with the traditional mean filter and wavelet threshold denoising, this method improves the PSNR and reduces the local peak relative error, and it can more efficiently remove the background noise and retain the target information.

吴强, 张锐. 基于局部异常因子的近地全天时星图小波去噪[J]. 光学学报, 2020, 40(8): 0810001. Qiang Wu, Rui Zhang. Wavelet Denoising of Near-Earth All-Day Star Map Based on Local Outlier Factor[J]. Acta Optica Sinica, 2020, 40(8): 0810001.

本文已被 6 篇论文引用
被引统计数据来源于中国光学期刊网
引用该论文: TXT   |   EndNote

相关论文

加载中...

关于本站 Cookie 的使用提示

中国光学期刊网使用基于 cookie 的技术来更好地为您提供各项服务,点击此处了解我们的隐私策略。 如您需继续使用本网站,请您授权我们使用本地 cookie 来保存部分信息。
全站搜索
您最值得信赖的光电行业旗舰网络服务平台!