动态摄像机图像传输信息优化提取仿真研究

Simulation Research on Optimal Extraction of Dynamic Camera Image Transmission Information

对动态图像的传输信息进行优化提取,能够有效提高动态模糊图像的清晰度。对图像传输信息的优化提取,需要依据特征点周围的局部图像信息,以矩阵的形式构造仿射不变特征描述因子,完成动态图像传输图像的优化提取。传统方法对基本特征成分的图像局部化进行约束,给出权值系数的稀疏分布,但忽略了构造仿射不变特征描述因子,导致提取精度较低。提出基于重要采样方法的动态图像传输信息优化提取方法。融合于重要采样原理将原始图像进行分块,利用样本图像对每个图像区域采样位置进行随机自适应采样,确定图像各个区域关键点的邻域,融合于小波理论依据特征点周围的局部图像信息,以矩的形式构造仿射不变特征描述因子,以此为依据完成对动态图像传输信息优化提取。实验结果表明,所提方法特征提取精度高,可以为提升动态图像的识别质量奠定了结实的基础。

An optimization extraction method for transmission information of dynamic camera image based on im- portance sampling method is proposed. The segmentation for original image integrated with the importance sampling method is carried out, sample image is used to carry out adaptive sampling for sampling location of each image region, and neighborhood of key point of each image region is confirmed. The description factor of invariant feature of affine in form of moment is built according to local image information around feature point integrated with wavelet theory. Thus, the. optimization extraction is completed. Simulation results show that the proposed method has high extraction precision. It can lay strong foundation for improving recognition quality of dynamic image.