• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.24 no.3
• /
• pp.384-390
• /
• 2020

The reverberation effect on the sound when producing movies or VR contents is a very important factor in the realism and liveliness. The reverberation time depending the space is recommended in a standard called RT60(Reverberation Time 60 dB). In this paper, we propose a scene recognition technique for automatic reverberation editing. To this end, we devised a classification model that independently trains color images and predicted depth images in the same model. Indoor scene classification is limited only by training color information because of the similarity of internal structure. Deep learning based depth information extraction technology is used to use spatial depth information. Based on RT60, 10 scene classes were constructed and model training and evaluation were conducted. Finally, the proposed SCR + DNet (Scene Classification for Reverb + Depth Net) classifier achieves higher performance than conventional CNN classifiers with 92.4% accuracy.

• Korean Journal of Remote Sensing
• /
• v.27 no.2
• /
• pp.75-87
• /
• 2011

This paper presents a method for region-based building extraction from high resolution satellite images(HRSI) using integrated information of spectral and color invariant features without user intervention such as selecting training data sets. The purpose of this study is also to evaluate the effectiveness of the proposed method by applying to IKONOS and QuickBird images. Firstly, the image is segmented by the MSRG method. The vegetation and shadow regions are automatically detected and masked to facilitate the building extraction. Secondly, the region merging is performed for the masked image, which the integrated information of the spectral and color invariant features is used. Finally, the building regions are extracted using the shape feature for the merged regions. The boundaries of the extracted buildings are simplified using the generalization techniques to improve the completeness of the building extraction. The experimental results showed more than 80% accuracy for two study areas and the visually satisfactory results obtained. In conclusion, the proposed method has shown great potential for the building extraction from HRSI.

• Journal of Advanced Navigation Technology
• /
• v.27 no.3
• /
• pp.306-313
• /
• 2023

Realistic and graphics-based virtual reality content is based on 360-degree videos, and viewport extraction through the viewer's intention or automatic recommendation function is essential. This paper designs a viewport extraction system based on multiple object tracking in 360-degree videos and proposes a parallel computing structure necessary for multiple viewport extraction. The viewport extraction process in 360-degree videos is parallelized by composing pixel-wise threads, through 3D spherical surface coordinate transformation from ERP coordinates and 2D coordinate transformation of 3D spherical surface coordinates within the viewport. The proposed structure evaluated the computation time for up to 30 viewport extraction processes in aerial 360-degree video sequences and confirmed up to 5240 times acceleration compared to the CPU-based computation time proportional to the number of viewports. When using high-speed I/O or memory buffers that can reduce ERP frame I/O time, viewport extraction time can be further accelerated by 7.82 times. The proposed parallelized viewport extraction structure can be applied to simultaneous multi-access services for 360-degree videos or virtual reality contents and video summarization services for individual users.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.5 no.1
• /
• pp.13-19
• /
• 2012

In this paper, we propose the pulmonary nodule detection method based on hierarchical 3D block analysis. The proposed system consists of two main part. In the first part, we select the block which is need to analysis. In the second part, we analysis the selected blocks. We extract the shape based features of the object in the selected blocks. Support Vector Machine is applied to the extracted features to classify into nodules and non-nodules.

• The Transactions of the Korea Information Processing Society
• /
• v.7 no.11
• /
• pp.3592-3602
• /
• 2000

Image processing consist of image analysis and classification. The one is extracting of feature value in the image. The other is segimentationof image that have same properiv. A novel approach for the analysis and classification of tezture images based on statistical texture prunitive estraction are proposed. In this approach, feature vector extracting is based on stalisucal method using apatial dependence of grey level and use general lexture proerty. In is advantageous that not effiected on structure and type of lexture. These components describe the amount of roughness and softness of texture images Two leatures. Moment and Homogeneity, are componted from GLCM(gray level co-occurrence matrices) of the lexture promitive to charactenize statisical properties of the image. We show the successful experimental results by considerationof these two components fro the analysis and classificationto regular and irregular texture images.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.11 no.5
• /
• pp.419-425
• /
• 2001

A method to detect lots of porno documents on the internet is presented in this parer. The proposed method applies fuzzy inference mechanism to the conventional information retrieval techniques. First, several example sites on porno arc provided by users and then candidate words representing for porno documents are extracted from theme documents. In this process, lexical analysis and stemming are performed. Then, several values such as tole term frequency(TF), the document frequency(DF), and the Heuristic Information(HI) Is computed for each candidate word. Finally, fuzzy inference is performed with the above three values to weight candidate words. The weights of candidate words arc used to determine whether a liven site is sexual or not. From experiments on small test collection, the proposed method was shown useful to detect the sexual sites automatically.

• Proceedings of the Korean Information Science Society Conference
• /
• 2006.10b
• /
• pp.480-484
• /
• 2006

본 논문에서는 코드 영역을 분리하기 위한 전처리 과정 중 코드 추출에 적합한 자동 이진화 알고리즘을 제안하여, 반복과정을 제거하고 정확한 코드영역 추출로 인식률 및 속도를 향상 시켰다. 배경이 복잡한 이미지가 들어 올 경우 기존의 전역 평균 임계값이나 클래스간의 분산을 이용한 방법으로는 이미지 코드 영역을 찾아 낼 수 없었던 문제를 해결하기 위하여 이미지 코드 주변에 배경과 구분을 두기 위한 흰색 영역이 있다는 점을 착안, 상하좌우 방향 바깥쪽에서 안쪽으로 탐색하여 가장 밝은 값을 갖는 값을 찾아내고 찾아낸 그룹 중 가장 낮은 값을 임계값으로 선택하여 최적의 임계값을 찾아 내었고 이를 통해 복잡한 영상 내에서도 이미지 코드 영역을 찾아낼 수 있다. 제안된 이진화 알고리즘의 성능을 평가하기 위하여 2000장의 테스트 이미지에 적용한 결과, 기존의 이진화 알고리즘들 보다 정확성뿐만 아니라 속도 면에서도 우수한 것을 확인하였다.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2012.07a
• /
• pp.275-278
• /
• 2012

본 논문에서 2D-3D 자동 영상 변환을 위하여 2D 상으로부터 깊이 지도(depth map)을 생성하는 방법을 제안한다. 제안하는 방법은 보다 정확한 깊이 지도 생성을 위해 영상의 전경 깊이 지도(foreground depth map)와 배경 깊이 지도(background depth map)를 각각 생성 한 후 결합함으로써 보다 정확한 깊이 지도를 생성한다. 먼저, 전경 깊이 지도를 생성하기 위해서 라플라시안 피라미드(laplacian pyramid)를 이용하여 포커스/디포커스 깊이 지도(focus/defocus depth map)를 생성한다. 그리고 블록정합(block matching)을 통해 획득한 움직임 시차(motion parallax)를 이용하여 움직임 시차 깊이 지도를 생성한다. 포커스/디포커스 깊이 지도는 평탄영역(homogeneous region)에서 깊이 정보를 추출하지 못하고, 움직임 시차 깊이 지도는 움직임 시차가 발생하지 않는 영상에서 깊이 정보를 추출하지 못한다. 이들 깊이 지도를 결합함으로써 각 깊이 지도가 가지는 문제점을 해결하였다. 선형 원근감(linear perspective)와 선 추적(line tracing) 방법을 적용하여 배경깊이 지도를 생성한다. 이렇게 생성된 전경 깊이 지도와 배경 깊이 지도를 결합하여 보다 정확한 깊이 지도를 생성한다. 실험 결과, 제안하는 방법은 기존의 방법들에 비해 더 정확한 깊이 지도를 생성하는 것을 확인할 수 있었다.

• Journal of the Korea Society of Computer and Information
• /
• v.13 no.6
• /
• pp.1-12
• /
• 2008

This paper presents automatic moving object detection method using the rotating camera covering larger area with a single camera. The proposed method is based on the edge segment matching which robust to the dynamic environment with illumination change and background movement. The proposed algorithm presents an edge segment based background panorama image generation method minimizing the distortion due to image stitching, the background image generation method using Generalized Hough Transformation which can reliably register the current image to the panorama image overcoming the stitching distortions, the moving edge segment extraction method that overcome viewpoint difference and distortion. The experimental results show that the proposed method can detect correctly moving object under illumination change and camera vibration.

• The Transactions of the Korea Information Processing Society
• /
• v.6 no.4
• /
• pp.868-876
• /
• 1999

In this paper, an efficient automatic video parsing technique on MPEG-compressed video that is fundamental for content-based indexing is described. The proposed method detects scene changes, regardless of IPB picture composition. To detect abrupt changes, the difference measure based on the dc coefficient in I picture and the macroblock reference feature in P and B pictures are utilized. For gradual scene changes, we use the macroblock reference information in P and B pictures. the process of scene change detection can be efficiently handled by extracting necessary data without full decoding of MPEG sequence. The performance of the proposed algorithm is analyzed based on precision and recall. the experimental results verified the effectiveness of the method for detecting scene changes of various MPEG sequences.