Comparison of Image Smoothing Methods on Potholes Road Images

Before segmentation performed on potholes, road images is done first on the image smoothing. With the aim at the time of the next stage of the next process (segmentation) is expected to produce good accuracy. In this research, there will be a comparison of four smoothing methods consisting of lowpass filter, highpass filter, highboost filter and Gaussian filter. The four smoothing methods applied to twenty pothole road image data that have been preprocessed using a combination of brightness and contrast methods. Comparative measurements of the four methods use the Means Square Error (MSE) value and Peak Signal to Noise Ratio (PSNR). Test results obtained from this research is the average value of MSE lowpass filter 68.12, MSE highpass filter is 569.99, MSE highboost filter is 838.54 and MSE Gaussian filter is 34.68. The average value of PSNR lowpass filter is 66.60 dB, PSNR highpass filter is 47.36 dB, PSNR highboost filter is 44.77 dB, and PSNR Gaussian filter is 75.94 dB. The measurement results obtained by the Gaussian filter method which is the best method for smoothing on potholes road images, because it produces the greatest PSNR value, and the smallest MSE value.