摘要
针对缺帧环境下图像的遮挡问题,提出一种新的特征优化提取方法。建立缺帧环境下遮挡图像采集模型,在此基础上,利用正方形网格建立立体模型,实现图像采集。通过阈值降噪法完成对遮挡图像的降噪滤波处理。DCS-LBP算子通过梯度方向上变化的灰度值实现编码,效果不佳。通过GCS-LDP算子对其进行优化,将二者结合在一起,对缺帧环境下遮挡图像施行编码,获取最后的直方图;将该直方图看作遮挡图像的特征,完成遮挡图像特征的优化提取。实验结果表明,所提方法能够有效对缺帧环境下遮挡图像进行预处理;且识别效率高、计算量少。
A new method of feature extraction is proposed to solve the problem of image occlusion. In order to realize image acquisition,the image acquisition model is set up in the absence of frame. The noise reduction filtering of the occluded image is completed by threshold denoising method. The gray DCS-LBP operator by gradient direction change value encoding,poor effect,by optimizing the GCS-LDP operator,the two together on the lack of frame under the environment of the occluded image encoding implementation,the final histogram was got. The histogram as features of images,the occluded image feature extraction was optimized. The experimental results show that the proposed method can effectively process the occluded image in the frame less environment,and the recognition efficiency is high and the computation is less.
出处
《科学技术与工程》
北大核心
2017年第35期87-91,共5页
Science Technology and Engineering
基金
江苏省现代教育技术研究课题(2015-R-4585)资助
关键词
缺帧
遮挡图像
特征
优化
提取
missing frame
occluded image
feature
optimization
extraction