CAJ | 학술논문

为克服常用平滑滤波算法不能兼顾去噪和保持图像细节的不足,提出了加权有向平滑滤波算法.先根据相似者相容的原理得到待处理像素与各滤波模板的隶属关系,从而判断待处理像素是噪声点,是图像本身像素,还是图像背景像素,然后选用相应加权有向或无向平滑模板对待处理像素进行平滑处理.实验仿真结果表明,加权有向平滑滤波算法既能有效地滤除灰度图像中的噪声,又能很好保护图像的边缘和细节,弥补了常用空域平滑滤波算法不能兼顾去噪和保持图像细节的不足.
위극복상용평활려파산법불능겸고거조화보지도상세절적불족,제출료가권유향평활려파산법.선근거상사자상용적원리득도대처리상소여각려파모판적대속관계,종이판단대처리상소시조성점,시도상본신상소,환시도상배경상소,연후선용상응가권유향혹무향평활모판대대처리상소진행평활처리.실험방진결과표명,가권유향평활려파산법기능유효지려제회도도상중적조성,우능흔호보호도상적변연화세절,미보료상용공역평활려파산법불능겸고거조화보지도상세절적불족.
In order to overcome the shortage that the traditional smoothing algorithm can not keep gradation images'edges and details well as well as filters out noises,a weighted and directional smoothing algorithm is introduced.First to judge which templates the pending pixel should belong to according to compatibility and similitude principle,then using the corresponding weighted and directional template or nodireetional template to smooth the pending pixel.The simulation results show that the weighted and directional smoothing algorithm filters out noises as well as keeps medical gradation images'edges and details well.In this respect,the algorithm remedies the shortage of common smoothing algorithm.