Soil Image De-Noising using Hyper Wavelet Double Window Median Filter (HWDWM)

In soil classification digital images are playing a vital role to analyze the quality of soil and its features. It also plays an important position in the following fields such as monitoring traffic, geographic information system, astronomy and handwriting recognition. Some of the noise may occur during the capture of the image. It contains different types of noise. The visual quality of the image is affected by noise and also decreases the brightness of the image. Noise may create unwanted effects in images such as blurring, strips, damages of edges and unseen lines. It also has a bad impression and surrounding of the image. So it is necessary to reduce the noise and also it will improve the quality of the image. Removing the noise from soil is a very big challenge for researchers. This study proposes a Hyper Wavelet Double Window Median Filter (HWDWM) to remove the noise from the soil images and enhance the quality of the soil image. The proposed filter outperforms the Double windows based coupled window mean filter (DBCWMF) and Discrete Wavelet Transform(DWT) in terms of Peak Signal to Noise Ratio(PSNR) and Mean Square Error(MSE).