• The Transactions of the Korea Information Processing Society
• /
• v.13 no.7
• /
• pp.335-347
• /
• 2024

Open-vocabulary 3D point cloud instance segmentation (OV-3DIS) is a challenging visual task to segment a 3D scene point cloud into object instances of both base and novel classes. In this paper, we propose a novel model Open3DME for OV-3DIS to address important design issues and overcome limitations of the existing approaches. First, in order to improve the quality of class-agnostic 3D masks, our model makes use of T3DIS, an advanced Transformer-based 3D point cloud instance segmentation model, as mask proposal module. Second, in order to obtain semantically text-aligned visual features of each point cloud segment, our model extracts both 2D and 3D features from the point cloud and the corresponding multi-view RGB images by using pretrained CLIP and OpenSeg encoders respectively. Last, to effectively make use of both 2D and 3D visual features of each point cloud segment during label assignment, our model adopts a unique feature ensemble method. To validate our model, we conducted both quantitative and qualitative experiments on ScanNet-V2 benchmark dataset, demonstrating significant performance gains.

• Journal of IKEEE
• /
• v.2 no.2 s.3
• /
• pp.185-193
• /
• 1998

Fractal-based monochrome image coding method can be easily applied for color image compression by splitting the color image into different primary spectral channels such as RGB, YIQ or $YC_bC_r$, and encoding each channel independently According to this method, it needs to repeat the fractal coding for each channel, so it have the problem of encoding time. In this paper, a fractal-based coder for color still image is proposed which features the enhancement of compression rate and the reduction of coding time. As the result of the experiment where the proposed algorithm is applied far color images, the compression rate is enhanced by 28 : 1 above with average PSNR value $28{\sim}29[dB]$, do not lossless encoding process using JPEG. And the encoding time is reduced by maximum 11.5 %.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.17 no.12
• /
• pp.2927-2936
• /
• 2013

In this paper, we propose a noise reduction and hole removal algorithm in order to improve the quality of depth images when they are used for creating 3D contents. In the proposed algorithm, the depth image and the corresponding color image are both used. First, an intensity image is generated by converting the RGB color space into the HSI color space. By estimating the difference of distance and depth between reference and neighbor pixels from the depth image and difference of intensity values from the color image, they are used to remove noise in the proposed algorithm. Then, the proposed hole filling method fills the detected holes with the difference of euclidean distance and intensity values between reference and neighbor pixels from the color image. Finally, we apply a parallel structure of GPGPU to the proposed algorithm to speed-up its processing time for real-time applications. The experimental results show that the proposed algorithm performs better than other conventional algorithms. Especially, the proposed algorithm is more effective in reducing edge blurring effect and removing noise and holes.

• Journal of Korean Academy of Oral Health
• /
• v.41 no.4
• /
• pp.290-295
• /
• 2017

Objectives: Dental plaque emits red fluorescence under a visible blue light near the ultra-violet end of the light spectrum. The fluorescence characteristics of each microorganism have been reported in several studies. The aim of this study was to evaluate changes in red fluorescence of oral microorganisms that is affected by blood in the culture media. Methods: The gram-positive Actinomyces naeslundii (AN, KCTC 5525) and Lactobacillus casei (LC, KCTC 3109) and gram negative Prevotella intermedia (PI, KCTC 3692) that are known to emit red fluorescence were used in this study. Each bacterium was activated in broth and cultivated in different agar media at $37^{\circ}C$ for 7 days. Tryptic soy agar with hemin and vitamin $K_3$ (TSA), TSA with sheep blood (TSAB), basal medium mucin (BMM) medium, and BMM with sheep blood (BMMB) were used in this study. Fluorescence due to bacterial growth was observed under 405-nm wavelength blue light using the quantitative light-induced fluorescence-digital (QLF-D) device. The red, green, and blue fluorescence values of colonies were obtained using image-analysis software and the red to green ratio (R/G value) and red to total RGB ratio (R/RGB value) were calculated for quantitative comparison. Results: The QLF-D images of the AN, LC, and PI colonies showed red fluorescence in all media, but the fluorescence of all bacteria was reduced in TSA and BMM media, compared with in TSAB and BMMB media. Both the R/G and the R/RGB values of all bacteria were significantly reduced in growth media without blood (P<0.001). Conclusions: Based on this in vitro study, it can be concluded that red fluorescence of oral bacteria can be affected by growth components, especially blood. Blood-containing medium could be a significant factor influencing red fluorescence of oral bacteria. It can be further hypothesized that bleeding in the oral cavity can increase the red fluorescence of dental plaque.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2000.10a
• /
• pp.522-527
• /
• 2000

본 논문은 그레이이미지를 컬러이미지로 변환하는 의사칼라이미지 향상에 관한 연구이다. 적용기법은 그레이이미지의 시간영역신호를 주파수영역으로 변환하고 LPF, BPF, HPF 2차원필터를 통과시켜 시간영역으로 역 변환한 후 각각에 대해 히스토그램 평활화하여 RGB 신호를 구하는 과정에서 의사 칼라를 얻는다. 실험 결과 필터주파수가 증가함에 따라 PSNR도 증가하고 낮게 설정하면 경계부분의 화질이 좋아진다. 2차원 주파수필터의 사용은 칼라 변환 시 강조하고자 하는 부분을 보다 자유롭게 변환할 수 있고 다양한 컬러 이미지로 향상되어 영상분석에 효과가 있다. 의사칼라의 변환에 의한 식별 인식능력은 9dB 정도 향상된다.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.26 no.3D
• /
• pp.519-525
• /
• 2006

The traditional area-based matching or efficient matching methods using epipolar geometry and height restriction of stereo images, which have a confined search space for image matching, have still some disadvantages such as mismatching and timeconsuming, especially in the dense metropolitan city that very high and similar buildings exist. To solve these problems, a new image matching method through building recognition has been presented. This paper described building recognition in color stereo images using edge and color information as a elementary study of new matching scheme. We introduce the modified Hausdorff distance for using edge information, and the modified color indexing with 3-D RGB histogram for using color information. Color information or edge information alone is not enough to find conjugate building pairs. For edge information only, building recognition rate shows 46.5%, for color information only, 7.1%. However, building recognition rate distinctly increase 78.5% when both information are combined.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.51 no.4
• /
• pp.173-180
• /
• 2014

Compressive sensing is a new signal acquisition paradigm that enables sparse/compressible signal to be sampled under the Nyquist-rate. To fully benefit from its much simplified acquisition process, huge efforts have been made on improving the performance of compressive sensing recovery. However, concerning color images, compressive sensing recovery lacks in addressing image characteristics like energy distribution or human visual system. In order to overcome the problem, this paper proposes a new group-sparsity hard thresholding process by preserving some RGB-grouped coefficients important in both terms of energy and perceptual sensitivity. Moreover, a smoothed group-sparsity iterative hard thresholding algorithm for compressive sensing of color images is proposed by incorporating a frame-based filter with group-sparsity hard thresholding process. In this way, our proposed method not only pursues sparsity of image in transform domain but also pursues smoothness of image in spatial domain. Experimental results show average PSNR gains up to 2.7dB over the state-of-the-art group-sparsity smoothed recovery method.

• Journal of Korea Multimedia Society
• /
• v.9 no.4
• /
• pp.434-442
• /
• 2006

This paper proposes a color-matching 3D look-up table that simplifies the complex color-matching procedure between a monitor and a mobile display device, where the image colors are processed in a device-independent color space, such as CIEXYZ or CIELAB, and gamut mapping performed to compensate the gamut difference. The transform from a device-dependent RGB color space to a device-independent color space is implemented by performing display characterization. The mobile LCD characterization error using the S-curve model is larger than the tolerance error since the mobile LCD has the channel-chromaticity-inconstancy and channel-dependence characteristics. In this paper we reduced the characterization error using the electro-optical transfer functions of X, Y, and Z value for R, G, B, C, M, Y, K components. Experimental results demonstrated that 64 ($4{\times}4{\times}4$) was the smallest size of color-matching look-up table that could produce an image with an acceptable reproduction error, based on a comparison of color-matched images resulting from the proposed color-matching look-up table and complex step-by-step color-matching procedures.

• Journal of KIISE:Computer Systems and Theory
• /
• v.35 no.7
• /
• pp.340-353
• /
• 2008

A computational color palette extraction model is introduced to describe paint brush objectively and efficiently. In this model, a color palette is defined as a minimum set of colors in which a painting can be displayed within error allowance and extracted by the two step processing of color grouping and major color extraction. The color grouping controls the resolution of colors adaptively and produces a basic color set of given painting images. The final palette is obtained from the basic color set by applying weighted k-means clustering algorithm. The extracted palettes from several famous painters are displayed in a 3-D color space to show the distinctive palette styles using RGB and CIE LAB color models individually. And the two experiments of painter classification and color transform of photographic image has been done to check the performance of the proposed method. The results shows the possibility that the proposed palette model can be a computational color analysis metric to describe the paint brush, and can be a color transform tool for computer graphics.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.17 no.1
• /
• pp.15-31
• /
• 2013

This paper proposes an intermediate image generation method for the viewer's view point by tracking the viewer's face, which is converted to a digital hologram. Its purpose is to increase the viewing angle of a digital hologram, which is gathering higher and higher interest these days. The method assumes that the image information for the leftmost and the rightmost view points within the viewing angle to be controlled are given. It uses a stereo-matching method between the leftmost and the rightmost depth images to obtain the pseudo-disparity increment per depth value. With this increment, the positional informations from both the leftmost view point and the rightmost view point are generated, which are blended to get the information at the wanted intermediate viewpoint. The occurrable dis-occlusion region in this case is defined and a inpainting method is proposed. The results from implementing and experimenting this method showed that the average image qualities of the generated depth and RGB image were 33.83[dB] and 29.5[dB], respectively, and the average execution time was 250[ms] per frame. Also, we propose a prototype system to service digital hologram interactively to the viewer by using the proposed intermediate view generation method. It includes the operations of data acquisition for the leftmost and the rightmost viewpoints, camera calibration and image rectification, intermediate view image generation, computer-generated hologram (CGH) generation, and reconstruction of the hologram image. This system is implemented in the LabView(R) environments, in which CGH generation and hologram image reconstruction are implemented with GPGPUs, while others are implemented in software. The implemented system showed the execution speed to process about 5 frames per second.