Apparatus and method for reducing noise of image sensor
AssigneeSamsung Electronics Co., Ltd.
InventorsYoung-jin Yoo, Seong-deok Lee, Hyun-wook Ok
Life of this patent
Estimating the term…
Abstract
An apparatus and method for reducing noise of an image sensor are provided. The apparatus includes a noise level setting module setting a noise level representing an image frame expressed as a component in a color space; a kernel generation module generating a kernel for filtering pixels that constitute the image frame based on the set noise level; and a filter application module convoluting the pixels using the generated kernel and normalizing the pixels using the convolution result. Since noise generated by the image sensor can be adaptively reduced, the quality of an output image can be improved.
Claims37
An apparatus for reducing noise of an image sensor, the apparatus comprising:
- a noise level setting module that sets a noise level representing an image frame expressed as a component in a color space;
- a kernel generation module that generates a kernel for filtering pixels constituting the image frame based on the noise level set by the noise level setting module;
- and
- a filter application module that convolutes the pixels using the kernel and normalizes the pixels based on a convolution result.
A method of reducing noise of an image sensor, the method comprising:
- setting a noise level that is representative of an image frame expressed as a component in a color space;
- generating a kernel that filters pixels that constitute the image frame based on the noise level set by the setting;
- and
- convoluting the pixels using the kernel and normalizing the pixels using a result of the convoluting.
A computer-readable recording medium on which a program code for executing a method of reducing noise of an image sensor is recorded, the method comprising:
- setting a noise level that is representative of an image frame expressed as a component in a color space;
- generating a kernel that filters pixels that constitute the image frame based on the noise level set by the setting;
- and
- convoluting the pixels using the generated kernel and normalizing the pixels using the convolution result.
Dependencies are inferred from the claim text.