Method and apparatus for detecting edge in image
AssigneeSamsung Electronics Co., Ltd.
InventorJin Gu HEO
Life of this patent
Estimating the term…
Abstract
A method and apparatus for detecting an image edge that detects an edge area from an input image is disclosed. The method of detecting the image edge includes removing noise of an input image, generating an intermediate image in which an edge area is accentuated based on the input image and the input image from which the noise is removed, determining a threshold value to be applied to the intermediate image through clustering the pixels of the intermediate image, and detecting an edge area from the intermediate image based on the threshold value.
Claims17
A method of detecting an image edge, the method comprising:
- removing noise of an input image;
- generating an intermediate image in which an edge area is accentuated based on the input image and the input image from which the noise is removed;
- determining a threshold value to be applied to the intermediate image through clustering pixels of the intermediate image;
- and
- detecting an edge area from the intermediate image, based on the threshold value.
A method of detecting an image edge, the method comprising:
- applying a weight determined based on a location of a pixel and a pixel value of an input image to pixels of the input image;
- removing noise of the input image to which the weight is applied, using a Gaussian distribution;
- generating an intermediate image in which an edge area is accentuated based on the input image to which the weight is applied and the input image from which the noise is removed;
- and
- detecting an edge area from the intermediate image, based on a distribution of a pixel value of the intermediate image.
An apparatus for detecting an image edge, the apparatus comprising:
- a noise removing unit to remove noise of an input image;
- an intermediate image generation unit to generate an intermediate image in which an edge area is accentuated, based on the input image and the input image from which the noise is removed;
- a threshold value determination unit to determine a threshold value to be applied to the intermediate image through clustering pixels of the intermediate image;
- and
- an edge area detection unit to detect an edge area from the intermediate image based on the threshold value.
Dependencies are inferred from the claim text.