Abstract In modern society developed based on the digital communication devices, the uses of image devices such as digital TV, digital camera and smartphone is gradually increasing and the purpose of these image devices is to reproduce identical image with natural thorough display such as monitor an...
Abstract In modern society developed based on the digital communication devices, the uses of image devices such as digital TV, digital camera and smartphone is gradually increasing and the purpose of these image devices is to reproduce identical image with natural thorough display such as monitor and portable device. But, in the process such as transmission, acquisition and storage, the unintended noises generated by elemental cause of inner and outer system and inequable illumination are added in image and so that these images have different with natural image because image quality is deteriorated. Hereupon, interest of image noise reduction technology to enhance image quality, by removing noise which irregularly causes image degradedness is growing. Depending on the cause of the noise, which is added to the image, there are many kinds such as Salt & Pepper, AWGN (additive white Gaussian noise) and Speckle noise. And among these, noise which is commonly generated in natural phenomena is AWGN that noise distribution follows a Gaussian distribution. Consequently, the studies to remove AWGN compared to other noises is animately proceeding and various methods have been proposed. The most basic way to remove noise is method of weighted mask using relationship of the central pixel and the adjacent pixel, after the mask is extracted from original image in spatial domain. It is simple to implement these methods because they have specific rules. But, when the image is processed by typical noise removing technique in spatial domain such as MF(mean filter) and GF(Gaussian filter), image appears disadvantages of blur phenomena and so on, especially, if the edge distortion of the image, the image of the representation is insufficient. On the other hand, the edge of the image is one of the important elements consisting image in order to process image and is the rapidly changing part at the image because it has information such as location, magnitude and orientation. The various studies for detecting edge has even been going now. And the representative edge detection method is the method using masks in the spatial domain proposed by Sobel, Prewitt, Roberts and so on. Whereas the method using mask in the spatial domain is relatively simple and short to process image, implements same operation in all locations, because consists in the subtraction between central pixel and the adjacent pixels. So, when AWGN is added to image, edge detection properties is insufficient. Therefore, in this paper, to effectively detect edges in AWGN environment, modified edge detection algorithm, using adaptive weighted value depending on spatial distance of central pixel and expanding area of mask in the spatial is proposed.
Abstract In modern society developed based on the digital communication devices, the uses of image devices such as digital TV, digital camera and smartphone is gradually increasing and the purpose of these image devices is to reproduce identical image with natural thorough display such as monitor and portable device. But, in the process such as transmission, acquisition and storage, the unintended noises generated by elemental cause of inner and outer system and inequable illumination are added in image and so that these images have different with natural image because image quality is deteriorated. Hereupon, interest of image noise reduction technology to enhance image quality, by removing noise which irregularly causes image degradedness is growing. Depending on the cause of the noise, which is added to the image, there are many kinds such as Salt & Pepper, AWGN (additive white Gaussian noise) and Speckle noise. And among these, noise which is commonly generated in natural phenomena is AWGN that noise distribution follows a Gaussian distribution. Consequently, the studies to remove AWGN compared to other noises is animately proceeding and various methods have been proposed. The most basic way to remove noise is method of weighted mask using relationship of the central pixel and the adjacent pixel, after the mask is extracted from original image in spatial domain. It is simple to implement these methods because they have specific rules. But, when the image is processed by typical noise removing technique in spatial domain such as MF(mean filter) and GF(Gaussian filter), image appears disadvantages of blur phenomena and so on, especially, if the edge distortion of the image, the image of the representation is insufficient. On the other hand, the edge of the image is one of the important elements consisting image in order to process image and is the rapidly changing part at the image because it has information such as location, magnitude and orientation. The various studies for detecting edge has even been going now. And the representative edge detection method is the method using masks in the spatial domain proposed by Sobel, Prewitt, Roberts and so on. Whereas the method using mask in the spatial domain is relatively simple and short to process image, implements same operation in all locations, because consists in the subtraction between central pixel and the adjacent pixels. So, when AWGN is added to image, edge detection properties is insufficient. Therefore, in this paper, to effectively detect edges in AWGN environment, modified edge detection algorithm, using adaptive weighted value depending on spatial distance of central pixel and expanding area of mask in the spatial is proposed.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.