论文部分内容阅读
物联网系统由感知层、传输层、应用层构成。感知层的数据精准度是物联网系统的关键要素。由感知层的众多传感器检测到同一目标的大量多维图像数据需要降维和提取典型特征,用以提高图像的质量。选用Gabor小波并设计滤波器的时域和频域以实现高频分量细化。通过Gabor小波变换将多维粗糙图像数据经过数据预处理、特征提取、特征匹配最终融合为单维精准数据。
IoT system consists of the perception layer, transport layer, application layer. The data accuracy of the perception layer is a key element of the IoT system. A large number of multi-dimensional image data of the same target detected by many sensors of the perception layer need to reduce dimension and extract typical features to improve the quality of the image. Select Gabor wavelets and filter time-domain and frequency domain to achieve high-frequency component refinement. The multi-dimensional rough image data is preprocessed by Gabor wavelet transform, and feature extraction and feature matching are finally fused into single-dimensional accurate data.