Wavelet based image denoising using raspberry PI

Abstract

There has been a huge demand for effective image restoration techniques since the increase in the production of the digital movies and images. No matter how good cameras are, an image improvement is always desirable to extend the image property and view. In image processing, it is very important to obtain precise images. Low image quality is an obstacle for effective feature extraction. Therefore, there is a fundamental need of noise reduction from images. current technique in reducing noise is efficient however focus on pc or desktop based processing. The drawback is the relatively high computational cost. This modification is done to non-local means algorithm results in improved accuracy and computational performance. There are currently a number of imaging modalities that are used for study of image processing. The aim of image de-noising in image processing is to clear the unwanted noise from the noisy image and implement an effective image denoising algorithm with the help of the Python programming language. Improve the processing methods by implementation of Non-local means and wavelet hard thresholding. Raspberry Pi-based system is used to implement image denoising. In this research, Gaussian noise is used, because noise property similar to a normal distribution. Both method has been implemented in Raspberry pi environment using Python 2.7.9 with Open CV 3.1.1. By used on both methods there are different result seen from the output. Non-local means (NLM) produce smoothen image and less noise, while wavelet hard thresholding have less error compared to Non-local means. The average difference between the PSNR by the hard threshold method and the PSNR by the NLM method of the images taken as an example of the image Lena the difference in succession according to the use of sigma 25 is 1.4711, in image Pepper 2.2303, in image Boat 2.0481. In term of processing time NLM is faster than wavelet hard thresholding. Both algorithm perform well in Raspberry Pi