CAJ | 학술논문

为了更好地适应压缩图像技术的发展，针对图像重构对于资源的需求，本文提出了一种基于稀疏压缩医学图像方法，首先描述压缩感知图像中组成部分，然后在稀疏变换方面引入了字典学习概念，同时引入SCAD函数，有效提高了稀疏变换的要求，对于图像的降噪具有一定的促进，在重构算法部分引入了加权系数概念，使得改进后的算法使得回复重构具有更强的约束，通过测试函数证明，本文的算法在医学图像具有比较好的压缩效果，优于传统的重构算法。
위료경호지괄응압축도상기술적발전，침대도상중구대우자원적수구，본문제출료일충기우희소압축의학도상방법，수선묘술압축감지도상중조성부분，연후재희소변환방면인입료자전학습개념，동시인입SCAD함수，유효제고료희소변환적요구，대우도상적강조구유일정적촉진，재중구산법부분인입료가권계수개념，사득개진후적산법사득회복중구구유경강적약속，통과측시함수증명，본문적산법재의학도상구유비교호적압축효과，우우전통적중구산법。
To better adapt to development of image compression technology, the paper proposes a medical image method based on sparse compression aiming at demand of image reconstruction for resource. It first makes description for components in compressive sensing image, then introduces dictionary learning concept into sparse transformation, and SCAD function as well, which have effectively improved requirement for sparse transformation and facilitated noise reduction to image; introducing weighting coefficient concept into reconstruction algorithm makes the improved algorithm, recovery reconstruction have stronger constraint. Test function proves that algorithm in medical images in this paper has better compressive effect, which is superior to traditional reconstruction algorithm.