公开(公告)号：KR102028427B1

公开(公告)日：2019-10-04

申请号：KR1020170147106

申请日：2017-11-07

Applicant: 주식회사 코렌텍 ,  전남대학교산학협력단

Inventor： 이정길 ,  이재원

IPC: A61F2/44 ,  A61F2/30

公开(公告)号：KR101634724B1

公开(公告)日：2016-06-30

申请号：KR1020150016483

申请日：2015-02-03

Applicant: 전남대학교산학협력단

Inventor： 홍성훈 ,  이재원

IPC: G06T5/00

CPC classification number: G06T5/009

Abstract: 본발명은밝기변환곡선생성및 선택을이용한적응적야간영상의밝기개선장치에관한것으로서, 야간영상과같이가시성이낮은어두운영상을실시간으로영상의특성에맞춰적응적으로가시성이향상된영상으로제공하는장치를제공함에목적이있다. 이러한목적을달성하기위한본 발명은, 다수개의밝기변환곡선들을생성하여저장하여곡선생성부; 입력영상의밝기분포와, 상기곡선생성부를통해생성된다수의밝기변환곡선들의기울기정보를이용하여입력영상에적합한밝기변환곡선을선택하는곡선선택부; 상기곡선선택부를통해선택된밝기변환곡선선택번호에해당하는밝기변환곡선을입력영상에적용하여밝기가개선된영상을출력하는밝기변환처리부; 및상기밝기변환처리부를통해밝기가개선된영상에대하여, 후처리를수행하여최종적으로가시성이향상된영상을출력하는후처리부; 를포함한다.

Abstract translation: 本发明涉及一种用于通过使用亮度转换曲线的生成和选择来自适应地提高夜间图像的亮度的装置，该亮度转换曲线可以实际地自适应地提高诸如夜间图像的低能见度的暗图像的亮度 时间根据图像的特点为图像提供改进的可见性。 本发明的装置包括：一个产生和存储多个亮度转换曲线的曲线产生单元; 曲线选择单元，用于通过使用输入图像的亮度分布和亮度转换曲线的斜率信息来选择适合于输入图像的亮度转换曲线; 亮度转换处理单元，用于通过向输入图像应用与由曲线选择单元选择的亮度转换曲线选择数对应的亮度转换曲线，输出具有改善的亮度的图像; 以及后处理单元，用于通过对由亮度转换处理单元改善亮度的图像执行后处理，最终输出具有改善的可视性的图像。

公开(公告)号：KR101384629B1

公开(公告)日：2014-04-11

申请号：KR1020120110000

申请日：2012-10-04

Applicant: 전남대학교산학협력단

Inventor： 이재원 ,  홍성훈

IPC: G06T5/00

CPC classification number: G06T5/007 ,  G06T5/10

Abstract: The present invention relates to a method for improving illumination intensity of an image and a computer-readable medium capable of being read with a computer in which an illumination improvement program performing the same is stored. More specifically, the present invention can obtain a consistent illumination improvement effect without complicated calculation by multiplying a frequency element of a low pass filter, which is comprised of cutoff filters including different weighted values according to a cutoff frequency, by the frequency element of a luminance signal of an input image. [Reference numerals] (10) Input image; (113) Weighted value calculation member; (180) Saturation improvement member; (190) Image separation member; (191) Image restoration member; (20) Output image

Abstract translation: 本发明涉及一种用于提高图像的照明强度的方法和能够用其中存储执行照明改进程序的计算机读取的计算机可读介质。 更具体地说，本发明可以通过将由截止滤波器构成的低通滤波器的频率元素，根据截止频率乘以不同的加权值，将亮度的频率元素乘以亮度 输入图像的信号。 （附图标记）（10）输入图像; （113）加权值计算成员; （180）饱和度改善部件; （190）图像分离部件; （191）图像恢复成员; （20）输出图像

公开(公告)号：KR1020140109217A

公开(公告)日：2014-09-15

申请号：KR1020130079499

申请日：2013-07-08

Applicant: 전남대학교산학협력단

Inventor： 홍성훈 ,  이재원

IPC: G06T5/00 ,  H04N5/217 ,  H04N5/57

CPC classification number: G06T5/40 ,  H04N5/217 ,  H04N5/57

Abstract: The present invention relates to a method and an apparatus for contrast enhancement of an image and, more specifically, to a method and an apparatus for contrast enhancement of an image using histogram compression equalization which transforms a histogram through a differential compression depending on size of the histogram; prevents distortions, such as supersaturation phenomenon, induced by excessive changes in brightness during quality enhancement of an image by equalizing the transformed histogram; and enhances contrast while maintaining the average brightness.

Abstract translation: 本发明涉及一种用于图像对比度增强的方法和装置，更具体地，涉及一种使用直方图压缩均衡的图像对比度增强的方法和装置，其通过差分压缩来变换直方图，取决于图像的大小 直方图; 通过均衡转换后的直方图，防止在图像质量增强期间通过亮度的过度变化而引起的过饱和现象; 并且在保持平均亮度的同时增强对比度。

公开(公告)号：KR101337423B1

公开(公告)日：2013-12-06

申请号：KR1020120103244

申请日：2012-09-18

Applicant: 전남대학교산학협력단

Inventor： 이재원 ,  홍성훈

IPC: H04N13/00 ,  G06T7/20

CPC classification number: G06T7/248 ,  G06T7/215 ,  G06T7/536

Abstract: The present invention relates to a moving object detecting and tracking method, more particularly, to a moving object detecting and tracking method using depth information and motion information, the method comprising: detecting a plurality of moving objects in real time with a lower complexity by using depth information and motion information of an input image; and updating and tracking an area of the moving object which is detected in the previous time. [Reference numerals] (AA) Start;(S1000) Input a current image;(S2000) Calculate frame different information;(S3000) Is the number of motion pixels greater than or equal to a threshold number?;(S4000) Compute depth projection information of the current image;(S5000) Divide and label an image object;(S6000) Is there a moving object?;(S7000) Is there a previous moving object?;(S8000) Register the state of the previous moving object as temporary disappearance or complete disappearance in a database;(S9000) Detect and track a moving object area

Abstract translation: 本发明涉及一种运动物体检测跟踪方法，更具体地涉及一种使用深度信息和运动信息的运动物体检测和跟踪方法，该方法包括：通过使用较低的复杂度实时检测多个运动物体 输入图像的深度信息和运动信息; 以及更新和跟踪在先前时间检测到的移动物体的区域。 （参考号）（AA）开始;（S1000）输入当前图像;（S2000）计算帧不同信息;（S3000）运动像素数大于或等于阈值数目;（S4000）计算深度投影 （S5000）对图像对象进行划分和标注（S6000）是否存在移动对象？（S7000）是否存在先前的移动对象？（S8000）将先前的移动对象的状态注册为临时 在数据库中消失或完全消失;（S9000）检测和跟踪移动物体区域