-
公开(公告)号：KR101418524B1
公开(公告)日：2014-08-06
申请号：KR1020130045330
申请日：2013-04-24
Applicant: 전자부품연구원
Abstract: A hardware apparatus and a method for creating an integral image are disclosed. The hardware apparatus calculating a feature point and descriptor of an image, includes: an image fading unit which reduces the maximum bit number into a predefined bit number by fading an entire inputted black and white image; an integral image creating unit which sequentially calculates pixel values of the black and white image outputted from the image fading unit to buffer, as an offset, a pixel value accumulated up to a previous line, and stores, as a differential value, a pixel value accumulated from a first pixel of a following line; and a first integral value acquisition unit which calculates a pixel value of a first integral image required to extract the feature point using the offset and the differential value.
Abstract translation: 公开了一种用于创建整体图像的硬件装置和方法。 计算图像的特征点和描述符的硬件装置包括：图像衰落单元，通过使整个输入的黑白图像衰落而将最大比特数减少到预定比特数; 积分图像生成单元，其顺序地计算从图像衰落单元输出的黑白图像的像素值，作为偏移量缓存累积到上一行的像素值，并将作为差分值的像素值 从下一行的第一像素累积; 以及第一积分值获取单元，其使用偏移量和差分值来计算提取特征点所需的第一积分图像的像素值。
-
公开(公告)号：KR101345127B1
公开(公告)日：2013-12-26
申请号：KR1020120156896
申请日：2012-12-28
Applicant: 전자부품연구원
IPC: H04N19/00
Abstract: Provided are a butterfly operation method for reducing computational complexity by operation skip and an HEVC system adopting the same. A butterfly operation method according to an embodiment of the present invention analyzes an input TPU and skips a butterfly operation for the TPU. As a result, computational complexity can be reduced by skipping transform/inverse transform operation itself through the analysis of values input into a butterfly structure in a transform/inverse transform process with a large amount of computational complexity in a HEVC system.
Abstract translation: 提供了一种通过操作跳过降低计算复杂度的蝶形运算方法和采用该方法的HEVC系统。 根据本发明实施例的蝶形运算方法分析输入TPU并跳过TPU的蝶形运算。 结果，通过在HEVC系统中具有大量计算复杂度的变换/逆变换处理中的输入到蝶形结构中的值的分析，可以通过跳过变换/逆变换操作本身来减少计算复杂度。
-
公开(公告)号：KR1020130123891A
公开(公告)日：2013-11-13
申请号：KR1020120047386
申请日：2012-05-04
Applicant: 전자부품연구원
CPC classification number: G06T3/0093 , H04N13/106
Abstract: Provided are a 3D warping method for suppressing the generation of holes and an image processing apparatus using same. The image processing method according to the embodiment of the present invention comprises dividing an image into multiple regions and performing 3D warping of the image by units of regions. Thus, holes cannot be generated in 3D warping, or the generation of holes can be suppressed to the maximum even when the holes occur such that a 3D-warped RGB image has correct pixel values and a 3D-warped Z-image has correct depth information. [Reference numerals] (110) RGB camera;(120) Z-camera;(131) Unit for correcting and arranging the RGB-image;(132) Edge detecting unit;(133) Joint extracting unit;(134) RGB-image dividing unit;(135) RGB-image 3D-warping unit;(136) Z-image correcting/arranging unit;(137) Z-image 3D-warping unit;(138) Z-image inpainting unit;(139) Z-image up-sampling unit;(140) Image application unit
Abstract translation: 提供一种用于抑制孔的产生的3D翘曲方法和使用其的图像处理装置。 根据本发明的实施例的图像处理方法包括将图像划分为多个区域并且以区域为单位执行图像的3D翘曲。 因此，即使发生三维弯曲的RGB图像具有正确的像素值并且3D变形的Z图像具有正确的深度信息，也可以在3D翘曲中产生孔，或者可以将孔的产生抑制到最大 。 （110）RGB相机;（120）Z摄像机;（131）用于校正和排列RGB图像的单元;（132）边缘检测单元;（133）联合提取单元;（134）RGB图像 分割单元;（135）RGB图像3D变形单元;（136）Z图像校正/排列单元;（137）Z图像3D扭曲单元;（138）Z图像修复单元;（139）Z- 图像上采样单元;（140）图像应用单元
-
公开(公告)号：KR1020130004732A
公开(公告)日：2013-01-14
申请号：KR1020110065974
申请日：2011-07-04
Applicant: 전자부품연구원
CPC classification number: H04N1/6027 , G06K9/6215 , G06T7/90
Abstract: PURPOSE: A similarity measurement method and a recording medium thereof are provided to improve accuracy of correspondence by searching a correspondence point of the correspondence image through detection and matching of a feature amount. CONSTITUTION: A computer readable code searches multiple correspondence points among correspondence images(S110). The code forms multiple correspondence blocks including the correspondence points among the correspondence images. The code produces first similarity by comparing color distribution of the correspondence blocks(S120). The code forms an image overlapped area among the entire area of the correspondence image. The code produces second similarity by comparing color distribution of the image overlapped areas(S130). The code measures similarity of the correspondence image by granting a weighting factor to the first and the second similarity(S140). [Reference numerals] (AA) Start; (BB) End; (S110) Searching a correspondence point between correspondence images; (S120) Generating a correspondence block and calculating a first similarity; (S130) Generating an overlapped area and calculating a second similarity; (S140) Measuring similarity of the correspondence image by granting a weighting factor to the first and the second similarities
Abstract translation: 目的：提供相似性测量方法及其记录介质，以通过检测和匹配特征量来搜索对应图像的对应点来提高对应的精度。 构成：计算机可读代码搜索通信图像中的多个对应点（S110）。 代码形成包括对应图像中的对应点的多个对应块。 该代码通过比较对应块的颜色分布来产生第一相似性（S120）。 代码在对应图像的整个区域中形成图像重叠区域。 该代码通过比较图像重叠区域的颜色分布来产生第二相似性（S130）。 代码通过将加权因子授予第一和第二相似度来测量对应图像的相似度（S140）。 （附图标记）（AA）开始; （BB）结束; （S110）搜索通信图像之间的对应点; （S120）生成对应块并计算第一相似度; （S130）生成重叠区域并计算第二相似度; （S140）通过将加权因子赋予第一和第二相似度来测量对应图像的相似度
-
公开(公告)号：KR1020120133424A
公开(公告)日：2012-12-11
申请号：KR1020110052049
申请日：2011-05-31
Applicant: 전자부품연구원
IPC: G11B20/10
CPC classification number: G11B20/00992 , G11B20/10046 , G11B2020/10564 , H04R2430/01
Abstract: PURPOSE: A method and an apparatus for automatic loudness control are provided to convert audio sounds into multiple critical band audio sounds and to apply them to LD and to assign bands included in the sound band in regular sequence. CONSTITUTION: A multi threshold band converter(110) converts input multichannel audio sounds into multiband audio sounds. The multi threshold band converter converts multiband audio sounds into multi threshold band audio sounds. A threshold band LD output unit(120) produces LDs for multi threshold band audio sounds. A threshold band LD difference allocator additionally assigns channel LD difference calculated by a channel LD-difference producer to threshold bands of the sound band.
Abstract translation: 目的：提供一种用于自动响度控制的方法和装置，用于将音频声音转换为多个临界频带音频声音，并将其应用于LD，并按规律顺序分配声带中包含的频带。 构成：多阈值频带转换器（110）将输入的多声道音频声音转换为多频带音频声音。 多阈值频带转换器将多频带音频声音转换为多阈值音频声音。 阈值带LD输出单元（120）产生用于多阈值频带音频声音的LD。 阈值频带LD差值分配器另外将由信道LD-差分产生器计算的信道LD差分给声带的阈值频带。
-
Patent Agency Ranking
公开(公告)号：KR1020120133041A
公开(公告)日：2012-12-10
申请号：KR1020110051519
申请日：2011-05-30
Applicant: 전자부품연구원
CPC classification number: G06T7/002 , G06T7/85 , H04N13/239 , H04N13/296
Abstract: PURPOSE: A correction system for a stereo camera and a stereo image correction device thereof are provided to improve a distortion correction result of a stereo image by interlinking physical correction for a stereo camera with image processing correction for the stereo image. CONSTITUTION: A rig control module(300) controls an electric rig according to an inputted camera correction parameter. The rig control module performs the physical correction of a camera. A stereo image correction device(500) obtains the camera correction parameter through correction for a stereo image. The stereo image correction device outputs the obtained camera correction parameter through a camera control module(400). The camera control module displays the camera correction parameter inputted form the stereo image correction device on a screen or through the rig control module. [Reference numerals] (110) Left lens; (120) Right lens; (140) Right image sensor; (150) Left image signal processor; (160) Right image signal processor; (200) Electric rig; (300) Rig control module; (400) Camera control module; (500) Stereo image correction device; (510) Stereo color correction module; (520) Geometric correction module(correction tool); (530) 3D depth information module; (600) Multiplexer; (AA) Corrected value
Abstract translation: 目的：提供一种用于立体相机及其立体声图像校正装置的校正系统，用于通过立体相机的物理校正与立体声图像的图像处理校正相互连接来改善立体声图像的失真校正结果。 构成：钻机控制模块（300）根据输入的相机校正参数控制电动钻机。 钻机控制模块执行相机的物理校正。 立体声图像校正装置（500）通过校正立体图像来获得相机校正参数。 立体图像校正装置通过照相机控制模块（400）输出获得的照相机校正参数。 相机控制模块在屏幕上或通过钻机控制模块显示从立体图像校正装置输入的相机校正参数。 （附图标记）（110）左镜片; （120）右镜片; （140）右图像传感器; （150）左图像信号处理器; （160）右图像信号处理器; （200）电钻机; （300）钻机控制模块; （400）相机控制模块; （500）立体声图像校正装置; （510）立体声色彩校正模块; （520）几何校正模块（校正工具）; （530）3D深度信息模块; （600）多路复用器; （AA）更正值
-
公开(公告)号：KR1020120082738A
公开(公告)日：2012-07-24
申请号：KR1020110004198
申请日：2011-01-14
Applicant: 전자부품연구원
CPC classification number: H04S3/008 , H04S2400/03
Abstract: PURPOSE: An audio system having a multi-channel and a multi-track and an audio processing method thereof are provided to prevent sound deterioration of audio objects taking a major role by reducing only data size of the audio objects taking a minor role. CONSTITUTION: An audio system(100) having a multi-channel and a multi-track comprises a multichannel audio mixer(110), channel down-mixers(120-1-120-P), audio encoders(130-1-130-P), and an audio MUX(140). The multichannel audio mixer performs an MIX operation by dividing channels from a multichannel audio object-K to a multichannel audio object-N. The channel down-mixers channel-down mix inputted multichannel audio objects. The audio encoders encode the audio objects channel-down mixed from the channel down mixers, respectively. The audio MUX generates single multi-channel multi-track audio.
Abstract translation: 目的：提供一种具有多声道和多声道及其音频处理方法的音频系统，以通过减少仅起作用的音频对象的数据大小来防止音频对象的声音恶化起主要作用。 构成：具有多声道和多声道的音频系统（100）包括多声道音频混合器（110），频道下混频器（120-1-120-P），音频编码器（130-1-130- P）和音频MUX（140）。 多通道音频混合器通过将多声道音频对象-K的信道划分为多声道音频对象N来执行MIX操作。 通道下混频器通道降低混合输入的多声道音频对象。 音频编码器分别对通道下混频器混合的音频对象进行编码。 音频MUX生成单个多声道多轨道音频。
-
公开(公告)号：KR101114431B1
公开(公告)日：2012-02-24
申请号：KR1020100004774
申请日：2010-01-19
Applicant: 전자부품연구원
IPC: G11B20/10
Abstract: 오디오 생성장치가 개시된다. 본 발명에 따른 오디오 생성장치는 오디오 객체를 입력받아 압축하는 오디오 인코더와, 오디오 객체에 일괄적으로 적용될 제1 오디오 효과를 포함하는 주요 장면묘사정보 및 오디오 객체들에 개별적으로 적용될 제2 오디오 효과를 포함하는 세부 장면묘사정보를 포함하는 스트리밍 장면묘사정보를 부호화하는 묘사인코더 및 스트리밍 장면묘사정보와 오디오 객체들을 결합하여 오디오 프레임 단위의 시간에 따라 전송되는 오디오 비트스트림을 생성하는 패킷화부를 포함한다. 본 발명에 따르면, 객체별 효과뿐만 아니라 전체 신호에 효과를 주기 위해서 장면효과 정보를 포함하고 있으며 각 효과를 적용하는 시간을 설정할 수 있다. 또한, 기존의 장면 묘사 방법과 달리 서비스 스트림의 중간에 사용자가 서비스를 받더라도 모든 장면 묘사 정보를 분석할 수 있도록 주요 정보를 반복적으로 전송함으로써 방송 서비스에 적용가능하다는 장점이 있다.
-
公开(公告)号：KR1020110111032A
公开(公告)日：2011-10-10
申请号：KR1020100030408
申请日：2010-04-02
Applicant: 전자부품연구원
IPC: G11B20/10 , G11B27/031
CPC classification number: G10L19/008
Abstract: 본 발명은 실감 객체 오디오 재생 및 생성 장치에 관한 것으로, 본 발명의 일면에 따른 실감 객체 오디오 재생 장치는 입력되는 오디오 파일로부터 SD(Scene Description) 압축 데이터 및 객체 오디오 압축 데이터를 각각 분리하는 디포맷터부, SD 압축 데이터를 복호화하여 SD 정보(Scene Description Information)를 복원하는 SD 복호화부, 객체 오디오 압축 데이터를 복호화하여 복수 객체 각각의 오디오 신호인 객체 오디오 신호를 복원하는 객체 오디오 복호화부 및 SD 정보 중 각 객체 오디오 신호에 대응하는 객체별 SD 정보에 따라 객체별 오디오 효과를 객체 오디오 신호에 부가하여 각 객체 오디오 신호에 대응하는 실감 객체 오디오 신호를 생성하는 객체 오디오 효과부를 포함한다.
Abstract translation: 根据本发明的一个方面的现实目标音频再现设备包括：变形器目标音频再现设备，用于从输入音频文件中分离SD（场景描述）压缩数据和目标音频压缩数据; 目标音频解码单元，用于解码目标音频压缩数据并恢复作为多个目标中的每一个的音频信号的目标音频信号;以及SD音频解码单元，用于解码SD音频数据， 以及对象音频效果单元，用于根据与对象音频信号相对应的对象特定SD信息将对象音频信号添加到对象音频信号以生成对应于每个对象音频信号的感觉对象音频信号。
-
公开(公告)号：KR1020110078498A
公开(公告)日：2011-07-07
申请号：KR1020090135324
申请日：2009-12-31
Applicant: 전자부품연구원
IPC: H04N7/24
CPC classification number: H04N1/644 , H04N19/186 , H04N19/593
Abstract: PURPOSE: A method for encoding image using estimation of color space is provided to improve an image compression efficient. CONSTITUTION: A method for encoding image using estimation of color space comprises steps of: encoding a pixel in a first color space(S110); calculating a color space difference prediction value of a second color space based on the encoded first color space(S120); and encoding the differential data of a current pixel value and the calculated color space prediction value(S130).
Abstract translation: 目的：提供一种使用颜色空间估计对图像进行编码的方法，以提高图像压缩效率。 构成：使用颜色空间估计对图像编码的方法包括以下步骤：对第一颜色空间中的像素进行编码（S110）; 基于编码的第一颜色空间计算第二颜色空间的色空间差预测值（S120）; 对当前像素值的差分数据和计算出的色彩空间预测值进行编码（S130）。
-
-
-
-
-
-
-
-
-
Search Results
163
records
(took 0.039 seconds)
