COMPOSITION THE SPATIAL FILTERING METHODS FOR QUALITY IMAGE

Abstract

The article deals with methods of improving the quality of digital images distorted noise. The distortion noise appears as a blending in the original image interfering signal. The most common types of noise is Gaussian noise, impulse noise and the result of complex action. For noise reduction using different methods of filtering images. To exclude noise investigated spatial filtering methods such as averaging, median and frequency filtering. Spatial averaging smooths noise emissions and frequency filtering minimizes signal noise level in the image. Results of filters lead to loss of image sharpness. To improve image sharpness using approaches that exhibit significant variations in the intensity of pixels. These methods include the use of Laplacian, but independent of its use leads to suppression of pixels without changes in brightness. Therefore there is the feasibility of compositions filtering methods to significantly improve image quality. The results of modeling of single use filtering methods to halftone images of their composition. We give an assessment methods used by criteria standard deviation, peak signal to noise ratio and structural similarity criterion.