Abstract:
이미지 센서로부터 출력되는 베이어 포멧의 영상으로부터 분리된 적색영상 및 청색영상의 스폿크기를 녹색영상의 스폿크기 수준으로 정정함으로써 색수차를 보정할 수 있는 소형 카메라 모듈의 색수차 보정 방법에 적용되는 각 색상영상의 블러링 양 측정 방법에 관한 것이다. 본 발명은, 이미지 센서가 촬상한 흑백 라인 에지에 대한 베이어 포맷의 영상을 취득하는 단계; 상기 베이어 포맷의 영상으로부터 적색화소로 이루어진 적색영상, 녹색화소로 이루어진 녹색영상 및 청색화소로 이루어진 청색영상을 분리하는 단계; 상기 적색영상, 녹색영상 및 청색영상에 대한 에지 프로파일 곡선을 구하는 단계; 상기 적색영상, 녹색영상 및 청색영상에 대한 에지 프로파일 곡선을 미분하여 상기 적색영상, 녹색영상 및 청색영상의 블러링 곡선을 구하는 단계; 및 상기 적색영상, 녹색영상 및 청색영상에 대한 블러링 곡선을 표현하는 각각의 블러링 함수를 결정하고, 상기 각 블러링 함수로부터 상기 적색영상, 녹색영상 및 청색영상에 대한 블러링 양을 산출하는 단계를 포함하는 소형 카메라 모듈의 블러링 양 측정 방법을 제공한다. 카메라 모듈, 렌즈, 광학계, 색수차, 이미지센서, 베이어(Bayer), RGB
Abstract:
영상을 복수의 블럭으로 구분하는 단계; 상기 각 블럭에 포함된 화소의 화소값을 수정하기 위한 화소 수정값을 저장하는 팔레트 룩업 테이블을 작성하는 단계; 상기 각 블럭에 포함된 화소들을 색공간 상에서 높은 휘도를 갖는 화소로 이루어진 제1 클러스터 및 낮은 휘도를 갖는 화소로 이루어진 제2 클러스터로 클러스터링하는 단계; 상기 각 블럭에 포함된 화소들의 색공간 상에서의 산포에 따른 에지 강도값을 측정하는 단계; 상기 각 블럭의 상기 에지 강도값을 나타내는 제1 인덱스 및 상기 2 개의 클러스터 중심값과 상기 블럭내 각 화소와의 거리에 따라 결정되는 제2 인덱스를 이용하여, 상기 팔레트 룩업 테이블로부터 상기 각 블럭 내 전체 화소 각각에 대한 화소 수정값을 결정하는 단계; 및 상기 화소 수정값을 해당 화소에 적용하는 단계를 포함하는 영상의 컨트라스트 향상 방법이 개시된다. 영상, 컨트라스트, 선명도, 이미지 처리, 클러스터, 색공간
Abstract:
A method for enhancing the sharpness of a color image is provided to improve the chromatic aberration problem and the deterioration of an image generated when using a low cost of lens such as a plastic lens. A process for dividing color images comprising plural pixels into plural blocks is performed(S21). A process for dividing the pixels within the divided blocks into two clusters by clustering is performed(S22). Thereafter, a process for correcting the color tone value of pixels is performed(S23) so that the distance between pixels belongs to two clusters is increased. The processes are performed by a sharpness improvement member. The color image processed by a sharpness improvement algorithm through the sharpness improvement member is processed by a DCF filtering manner additionally through a DCF filtering member.
Abstract:
An image processing method for improving sharpness is provided to improve the sharpness of an image by correcting chromatic aberration caused by difference of spot sizes of RGB channel images generated in a Bayer image detected from an image sensor. By applying a Gaussian differential filter to each color channel of a Bayer image including RGB(Red, Green and Blue) channels outputted from an image sensor, a gradient image for each color channel is generated(S41). After one of the RGB channels is determined as a reference channel and the rest color channels are determined as correction channels, a peak present at a common position of gradient images of the reference and correction channels is detected(S42). A parameter of the detected peak is calculated(S43). By using the parameter, a blurred edge function of the reference and correction channels is obtained. By using the edge function, a chromatic aberration correction mask is configured(S44). By applying the chromatic aberration correction mask to the correction channel, a value of pixels of the correction channel is corrected(S45).
Abstract:
PURPOSE: A camera unit for a vehicle, method for displaying outside a vehicle, and a system for generating driving corridor markers are provided to achieve a driving corridor display function without an additional external ECU(Electronic Control Unit). CONSTITUTION: A camera unit for a vehicle comprises an image sensor(13), an interface controller(12), and an image processing unit(14). The image sensor takes a picture of the image outside a vehicle. The interface controller stores a control points set with respect to a steering angle. A plurality of nodes are created by calculating or loading a control location of a Bezier curve with respect to a steering angle requested from the stored control points set. The image processing unit receives a coordinate of the nodes from the interface controller, creates a driving corridor mark by connecting the adjacent nodes with a line segment, and overlaps the created mark with the photographed image by an image sensor. The image sensor, and the interface controller and the image processing unit are installed in one camera unit.
Abstract:
A contrast improvement method of an image is provided to improve contrast of the image while reducing ringing noise. An image is classified by a plurality of blocks. A palette look up table storing a pixel correction value for modifying a pixel value of a pixel included in each block is written(S22). The first cluster having high luminance on a color space of pixels included in each block and the second cluster having low luminance are clustered(S23). An edge strength value according to dissemination on a color space of the pixels included in each block is measured(S25).
Abstract:
본 발명은, 플라스틱 렌즈와 같이 저가의 렌즈를 사용할 때 발생하는 색수차, 영상 열화를 개선하고 컬러 영상의 품질을 향상시킬 수 있는 컬러 영상의 선명도 향상 방법에 관한 것이다. 본 발명은, 복수의 화소로 이루어진 컬러 영상을 복수의 블록으로 구분하는 단계; 상기 각 블록 내의 화소들을 클러스터링하여 두 개의 클러스터로 구분하는 단계; 및 색공간 상에서 상기 두 클러스터에 속한 화소들 사이의 거리가 더 멀어지도록 상기 화소들의 색상값을 수정하는 단계를 포함하는 컬러 영상의 선명도 향상 방법을 제공한다. 플라스틱 렌즈, 색수차, 선명도, 색공간, 클러스터, 클러스터링
Abstract:
A method for measuring an amount of blurring in a small-sized camera module is provided to measure the blurring of an image generated by an optical system from a camera module capture image in a general illumination environment easily. An image of a bayer format outputted from an image sensor is directly obtained(S41) in order to calculate blurring functions. The red image comprising red pixels, green image comprising green pixel and blue image comprising blue pixels are separated from the bayer format image(S42). Edge profile curves on the extracted red image, green image and blue image are calculated(S43). Thereafter, the edge profile curves relating to the red image, green image and blue image are differentiated so that blurring curves of the red image, green image and blue image are obtained(S44). Each blurring functions for representing the blurring curves relating to red image, green image and blue image are determined(S45).
Abstract:
An auto-focusing method for a camera is provided to perform the auto-focusing of a lens of the camera quickly and accurately without using a position sensor in order to adjust a position of the lens. The sharpness of a red channel of an image picked up through an image sensor and the sharpness of a blue channel are acquired and compared(S100). When the difference between the sharpness of the red channel and the sharpness of the blue channel is within a predetermined error range, a process is ended(S200). When the difference between the sharpness of the red channel and the sharpness of the blue channel is out of the predetermined error range, the process is performed. When the sharpness of the red channel is higher than the sharpness of the blue channel(S300), a lens is moved in an image sensor direction and the distance between the lens and the image sensor is reduced(S400). When the sharpness of the red channel is lower than the sharpness of the blue channel, the lens is moved in an opposite direction of the image sensor and the distance between the lens and the image sensor is increased(S500).