农业物联网数字图像传输—小波变换信号滤波处理研究

Research on Digital Image Transmission and Wavelet Transform Signal Filtering Processing in the Agricultural Internet of Things

在农业物联网数字图像传输过程中,经常存在信号杂波会产生噪声,这对图像的像素与质量产生一定影响,如边缘检测、影像分析与处理、机器视觉等方面,噪声的干扰会对图像处理的结果产生畸变与误差。因此,小波变换对于线性与非线性问题中的非平稳信号具有良好的分析能力,故成为信号滤波的常用工具。混合中值滤波与小波变换结合改进算法的原理即在实际信号范畴中检索到与小波函数空间匹配的最佳映射,其目的在于使得原有传输信号达到最佳复原。小波滤波可以近似逼近为一个低通滤波,只是相比普通低通滤波而言,它还能保持原有信号某些尺度与细节特点。

In the process of digital image transmission in agricultural internet of things, there is often a noise, which affects the quality of image. Such as edge detection, image analysis and processing, machine vision and other aspects, the noise of image processing results will produce distortion and error. Wavelet transform is a common tool for signal filtering because of its good ability to analyze non-stationary signals in linear and nonlinear problems. The principle of the combination of median filter and wavelet transform is the best mapping of the spatial matching of wavelet function in the actual signal category, which aims to make the original transmission signal reach the best restoration. In other words, the wavelet filter can be approximated as a low pass filter, but compared to the ordinary low pass filter, it can keep the original signal certain scale and detail features.