摘要
当前超分辨率数字图像特征提取及重构方法容易受到外界环境的干扰,导致重构结果不可靠,重构图像质量较低。为此,提出一种新的超分辨率数字图像特征提取方法,通过BRISK描述子对超分辨率数字图像特征进行提取,以提高重构图像质量。详细分析了重构约束的构建过程;在此基础上,通过低分辨率数字图像与平滑性求解获取高分辨率数字图像,从而实现超分辨率数字图像的重构。实验结果表明,采用所提的新的超分辨率数字图像特征提取及重构方法对图像进行重构,不仅匹配性能高,而且重构图像质量优、效果佳。
The current super resolution digital image feature extraction and reconstruction method are vulnerable to the interference of the external environment, the reconstruction result unreliable, reconstructed image quality is low. For this, a new method of super resolution digital image feature extraction, through your BRISK descriptor was put forward for super-resolution digital image features extraction, in order to improve the quality of reconstructed image. Reconstruction constraint of the build process is analyzed in detail, on this basis, through the low resolution digital images with smooth solution to obtain high resolution digital image, so as to realize the super-resolution reconstruction of digital image. Experimental results show that the proposed new super-resolution digital image feature extraction and reconstruction method for image reconstruction, not only matching performance is high, and the reconstructed image quality, good effect.
出处
《科学技术与工程》
北大核心
2017年第11期255-259,共5页
Science Technology and Engineering
基金
西藏民族大学青年学人培育计划项目成果(16MYQP05)资助
关键词
超分辨率
数字图像
特征提取
重构
super resolution digital image feature extraction refactoring
