CAJ | 학술논문

针对监控视频图像的特点，提出了一种基于时空联合的实时视频降噪算法。该算法通过结合多帧图像进行运动检测，自适应地区分图像的运动区域和静止区域，对静止区域采用时域加权均值滤波，对运动区域采用空域ANL滤波。实验结果表明，该算法由于准确地区分了图像的运动区域和静止区域，充分利用了视频的时域、空域信息，在不造成运动拖影的前提下，能够显著提高视频的信噪比和图像的主观质量，同时满足实时性要求。
침대감공시빈도상적특점，제출료일충기우시공연합적실시시빈강조산법。해산법통과결합다정도상진행운동검측，자괄응지구분도상적운동구역화정지구역，대정지구역채용시역가권균치려파，대운동구역채용공역ANL려파。실험결과표명，해산법유우준학지구분료도상적운동구역화정지구역，충분이용료시빈적시역、공역신식，재불조성운동타영적전제하，능구현저제고시빈적신조비화도상적주관질량，동시만족실시성요구。
An algorithm of video real-time denoising based on spatio-temporal combination is proposed for the video surveillance syetem. By motion detection to multi-frame images, this algorithm can adaptively distinguish the still regions and motion regions of video image. Temporal weighted average filter to the still regions and spatial ANL filter to the motion regions are used separately. Experimental results show that because of the the precise distinction of the still regions and motion regions,fully utilizing the spatio-temporal information of the video sequences,the proposed algorithm can significantly improve the signal-to-noise ratio and the subjective image quality without movement ghosting, while the real-time requirements could be well satisfied.