Performance Analysis of Signal Denoising and Remote Sensing Image Denoising in Photoelectric Wireless Sensor Networks Based on Matched Wavelet

摘要

In the process of remote sensing data acquisition by photoelectric sensors, the sensor signal is often distorted due to equipment, illumination and other factors. Therefore, an important goal of remote sensing data acquisition is to eliminate noise as much as possible. Firstly, a wireless network topology structure based on PAN3101 photoelectric sensor is constructed. With CC2530 as the main control chip, the software simulation and device selection of node system design and circuit design of photoelectric sensor are carried out. The knowledge of traditional wavelet analysis and filter is studied, and taking the abnormal signal of photoelectric sensor as the research object, a wavelet matched method based on the specified abnormal signal is proposed. This method combines the wavelet analysis with the filter, sets matching principles, and obtains the matched wavelet. The wavelet matching method is applied to detect the transmission quality of optoelectronic sensor signal and compare with the performance of Haar algorithm and DBn algorithm for image denoising. In the process of experiment, the method of wavelet matching is used for signal denoising and image denoising. In the experiment of signal denoising, the method of wavelet matching proposed has obvious advantages over the classical method of wavelet denoising in the MSE (Mean Square Error) and PSNR (Power Signal-to-Noise Ratio) indexes of image. In the experiment of image denoising, the algorithm proposed is applied to remote sensing image denoising. The noise of image transmitted by sensor is suppressed and the edge information of image is well preserved. It shows that this method has certain practical value in wireless network constructed by photoelectric sensor.