摘要:
An auto white balance method includes converting a color space of an input image from an RGB color space into the Commission International de I'Eclairage (CIE) L*a*b* color space, determining at least a portion of pixels of the input image, assumed that a reference white color is changed, in a range of a predetermined L*a*b* value as pixels to be used for estimating the reference white color, determining averages of an L* value, an a* value, a b* value of the determined pixels to be used for estimating the reference white color as a reference white color estimation value, and calculating a color gain to move the reference white color estimation value to a target value for a predetermined white balance.
摘要翻译:自动白平衡方法包括将来自RGB颜色空间的输入图像的颜色空间转换为国际照相机(CIE)L * a * b *颜色空间,确定输入图像的至少一部分像素 假设在作为用于估计参考白色的像素的预定L * a * b *值的范围内改变参考白色,确定L *值,a *值,ab *的平均值, 用于估计参考白色的确定像素的值作为参考白色估计值,以及计算颜色增益以将参考白色颜色估计值移动到预定白平衡的目标值。
摘要:
There is provided a method of performing auto white balance including: setting a plurality of sections of a whole range of a value of each of Y, Cb, and Cr in the YCbCr color space, according to a size of each value; calculating an accumulated Y, Cb, and Cr value of pixels included in each of the plurality of sections; determining effective sections according to the number of the pixels included in each of the plurality of sections; calculating a mean Y, Cb, and Cr value of pixels included in all of the effective sections by using the accumulated Y, Cb, and Cr value of each of the effective sections; calculating a mean RGB value by converting the mean Y, Cb, and Cr value into an RGB color space; and determining an RGB color gain of a Bayer image outputted from an image sensor by using the mean RGB value.
摘要:
There is provided a method of performing auto white balance including: setting a plurality of sections of a whole range of a value of each of Y, Cb, and Cr in the YCbCr color space, according to a size of each value; calculating an accumulated Y, Cb, and Cr value of pixels included in each of the plurality of sections; determining effective sections according to the number of the pixels included in each of the plurality of sections; calculating a mean Y, Cb, and Cr value of pixels included in all of the effective sections by using the accumulated Y, Cb, and Cr value of each of the effective sections; calculating a mean RGB value by converting the mean Y, Cb, and Cr value into an RGB color space; and determining an RGB color gain of a Bayer image outputted from an image sensor by using the mean RGB value.
摘要:
There is provided a method of grouping pixels in a 2D digital image that assigns labels to regions of objects to differentiate between objects separated in the input image and groups a region of an object having the same label into one group when pattern recognition and image process are performed on the 2D digital image. A method of grouping pixels in a 2d digital image according to an aspect of the invention may include: determining whether a target pixel in an input image is included in a labeling section; determining whether the target pixel is connected to a previous pixel of the target pixel and an upper pixel located adjacent to and above the target pixel to perform a labeling process, and renewing a label table while saving a labeling result in a line memory when the target pixel is included in the labeling section and is a valid pixel including object information; and grouping pixels of the input image using the labeling result and the renewed label table, and renewing a group.
摘要:
There is provided an apparatus for removing color noise including: a color interpolation unit performing color interpolation on a bayer pattern image output from an image sensor; a high-pass filter unit performing high-pass filtering on each of the pixels of the bayer pattern image to generate high-pass filtered values of each of the pixels; a high frequency region determining unit comparing pixel values of a target pixel, from which color noise is removed, and pixels adjacent to the target pixel with the high-pass filtered values to determine whether the target pixel is included in a high frequency region; and a color noise region determining unit using differences between color values of the target pixel interpolated by the color interpolation unit and determining whether the target pixel is included in a color noise region when it is determined that the target pixel is included in the high frequency region.
摘要:
Provided is an apparatus for improving the sharpness of an image. The apparatus includes an edge detector, an effective edge judging unit, an effective edge classifying unit, an isolated edge judging unit, and a 2-D HPF applying unit. The edge detector calculates an edge value using brightness relation with neighboring pixels with respect to each of pixels of an input image. The effective edge judging unit determines effective edge pixels. The effective edge classifying unit determines weak edge pixels and strong edge pixels. The isolated edge judging unit judges an isolated edge pixel included in an isolated edge. The 2-D HPF applying unit applies a first 2-D HPF to the weak edge pixels, and applies a second 2-D HPF to the strong edge pixels to generate edge strengthening values. The edge strengthening value applying unit adds the edge strengthening values to brightness values of the respective pixels of the input image.
摘要:
Provided are an apparatus for removing ringing noise and an apparatus for removing noise. The apparatuses includes a representative value setting block dividing pixel data in one block including a plurality of pixel data into two groups through a K-means algorithm and setting a representative value in each of the groups; a parameter generator block calculating intervals and approaches between all the pixel data in the block and a pair of the representative values; and a correction block correcting an inputted pixel data value using the intervals and the approaches.
摘要:
Provided are an apparatus for removing ringing noise and an apparatus for removing noise. The apparatuses includes a representative value setting block dividing pixel data in one block including a plurality of pixel data into two groups through a K-means algorithm and setting a representative value in each of the groups; a parameter generator block calculating intervals and approaches between all the pixel data in the block and a pair of the representative values; and a correction block correcting an inputted pixel data value using the intervals and the approaches.
摘要:
A Two-Dimensional (2D) image segmentation apparatus for segmenting pixels of a progressive input 2D image includes a group information storing unit storing information of pixel groups including a plurality of adjacent pixels; a pixel determining unit determining coordinates of an input pixel, and determining whether the input pixel is an effective pixel for segmentation; a group scanning unit scanning a adjacent pixel group disposed in a scan range preset from the effective pixel in the group storing unit when the input pixel is determined as the effective pixel by the pixel determining unit; and a group information updating unit updating information of a pixel group stored in the group information storing unit according to whether there is the scanned adjacent pixel group and a pixel group including an input pixel preceding the effective pixel. Moreover, a method is provided that removes red-eye using area information of a pixel group determined by the segmentation apparatus.
摘要:
A color interpolation apparatus includes: a threshold value determining unit that calculates a luminance value of each of a plurality of areas set based on each pixel of a bayer pattern image generated by an image sensor, and determines edge determining threshold values used for determining each edge of the set areas; an edge information calculating unit that calculates pixel value variations in multiple directions of horizontal, vertical, diagonal, and central directions of the corresponding set areas by using pixel values of pixels included in each set area, and calculates edge information regarding each set area by using the calculated multi-directional pixel value variations; an area determining unit that determines a type of an edge present in each set area by using the edge determining threshold values deterred by the threshold value determining unit and the edge information calculated by the edge information calculating unit; and a color interpolation unit that applies a color interpolation filter previously set according to the edge types to a corresponding set area according to the edge types determined for each region set by the area determining unit.