• Journal of the Korean Society for Industrial and Applied Mathematics
• /
• 제25권2호
• /
• pp.26-38
• /
• 2021

Image coloration refers to adding plausible colors to a grayscale image or video. Image coloration has been used in many modern fields, including restoring old photographs, as well as reducing the time spent painting cartoons. In this paper, a method is proposed for colorizing grayscale images using a convolutional neural network. We propose an encoder-decoder model, adapting FusionNet to our purpose. A proper loss function is defined instead of the MSE loss function to suit the purpose of coloring. The proposed model was verified using the ImageNet dataset. We quantitatively compared several colorization models with ours, using the peak signal-to-noise ratio (PSNR) metric. In addition, to qualitatively evaluate the results, our model was applied to images in the test dataset and compared to images applied to various other models. Finally, we applied our model to a selection of old black and white photographs.

• 대한전자공학회논문지SP
• /
• 제43권6호
• /
• pp.65-84
• /
• 2006

본 논문에서는 개선된 영상 생성 모델을 제시하고, 제시된 모델에 기반한 칼라 영상 향상을 제안한다. 제시된 영상 생성 모델에서는 입력 영상을 전역 조명 성분과 국부 조명 성분, 그리고 반사율 성분의 곱으로 표현한다. 제안된 칼라 영상 향상에서는 RGB 입력 칼라 영상을 HSV 칼라 영상으로 변환한 다음, 백색광 조명 상태라는 가정 하에 개선된 영상 생성 모델에 근거하여 V 성분 영상만을 향상한다. 전역 조명 성분은 입력 V 성분 영상에 유효 영역이 넓은 선형 저대역 필터를 적용하여 추정하고, 국부 조명 성분은 입력 V 성분 영상에서 추정된 전역 조명 성분이 제거된 영상에 유효 영역이 좁은 JND (just noticeable difference) 기반의 비선형 저대역 필터를 적용하여 추정한다. 그리고 반사율 성분은 입력 V 성분 영상에 추정된 전역 조명 성분과 국부 조명 성분을 나누어 추정한다. 이어서 이들 추정된 성분에 감마 수정을 각각 적용하고 그 결과들을 곱하여 출력 V 성분 영상을 얻은 다음 히스토그램 모델링을 적용하여 최종 출력 V 성분 영상을 얻는다. 마지막으로 최종 출력 V 성분 영상과 입력 H 성분 영상 및 S 성분 영상으로부터 출력 RGB 칼라 영상을 얻는다. 실험 결과 제안된 방법은 NASA 홈 페이지로부터 다운받은 칼라 영상과 MPEG-7 CCD 칼라 영상으로 구축한 시험 영상 데이터 베이스에 대하여 후광 효과가 거의 억제되고 색상 변화가 거의 없으면서 전역 대비와 국부 대비를 동시에 잘 증가시키는 것을 확인하였다.

• 한국HCI학회:학술대회논문집
• /
• 한국HCI학회 2006년도 학술대회 1부
• /
• pp.772-779
• /
• 2006

The new model proposed in this paper is the hidden Markov mesh model or the 2D HMM with the causality of top-down and left-right direction. With the addition of the causality constraint, two algorithms for the evaluation of a model and the maximum likelihood estimation of model parameters have been developed theoretically which are based on the forward-backward algorithm. It is a more natural extension of the 1D HMM than other 2D models. The proposed method will provide a useful way of modeling highly variable image patterns such as offline cursive characters.

• 한국방송∙미디어공학회:학술대회논문집
• /
• 한국방송공학회 1999년도 KOBA 방송기술 워크샵 KOBA Broadcasting Technology Workshop
• /
• pp.133-140
• /
• 1999

In this paper, we propose a method of image-based three-dimensional modeling for realistic facial expression. In the proposed method, real human facial images are used to deform a generic three-dimensional mesh model and the deformed model is animated to generate facial expression animation. First, we take several pictures of the same person from several view angles. Then we project a three-dimensional face model onto the plane of each facial image and match the projected model with each image. The results are combined to generate a deformed three-dimensional model. We use the feature-based image metamorphosis to match the projected models with images. We then create a synthetic image from the two-dimensional images of a specific person's face. This synthetic image is texture-mapped to the cylindrical projection of the three-dimensional model. We also propose a muscle-based animation technique to generate realistic facial expression animations. This method facilitates the control of the animation. lastly, we show the animation results of the six represenative facial expressions.

• 대한원격탐사학회:학술대회논문집
• /
• 대한원격탐사학회 2002년도 Proceedings of International Symposium on Remote Sensing
• /
• pp.75-80
• /
• 2002

The mainly used technique to rectify satellite images with distortion is to develop a mathematical relationship between the pixel coordinates on the image and the corresponding points on the ground. By defining the relationship between two coordinate systems, a polynomial model is designed and various linear transformations are used. These GCP based geometric correction has performed overall plane to plane mapping. In the overall plane mapping, overall structure of a scene is considered, but local variation is discarded. The highly variant height of region is resampled with distortion in the rectified image. To solve this problem this paper proposed the TIN-based rectification on a satellite image. The TIN based rectification is good to correct local distortion, but insufficient to reflect overall structure of one scene. So, this paper shows the experimental result and the analysis of each rectification model. It also describes the relationship GCP distribution and rectification model. We can choose a geometric correction model as the structural characteristic of a satellite image and the acquired GCP distribution.

• 미래기술융합논문지
• /
• 제2권3호
• /
• pp.49-58
• /
• 2023

본 연구는 브랜드 이미지와 모델 이미지 일치여부에 따른 화장품 광고효과와 브랜드 이미지와 광고 카피 간을 조화롭게 지각하느냐 여부에 따른 화장품 광고효과를 알아봄으로써, 화장품 광고에서 맥락효과를 검증하고자 하였다. 이를 위해 브랜드 가치유형(3)×광고카피유형(3) 요인설계를 이용하여 자료를 수집하였다. 연구결과 첫째, 화장품 브랜드 이미지와 모델 이미지의 일치여부에 따른 광고효과를 확인한 결과, 모델의 이미지와 브랜드의 이미지가 일치할 때 광고태도와 구매의도가 모두 유의미하게 높은 것으로 나타났다. 둘째, 화장품 브랜드 이미지와 카피유형의 일치도 지각여부에 따라 광고효과에서 차이가 있는 가를 확인하였다. 그 결과 화장품 브랜드 이미지와 카피유형이 일치한다고 지각하는 소비자들이 불일치한다고 지각하는 소비자에 비해 광고태도와 구매의도가 모두 유의미하게 높은 것으로 나타났다. 향후 화장품 광고 카피 전략을 세울 때맥락효과를 접목하여 카피전략을 세워야 하는가에 대한 타당성을 제공해 줄 것으로 기대된다.

• 소프트웨어공학소사이어티 논문지
• /
• 제29권1호
• /
• pp.13-21
• /
• 2020

CNN 모델이 이미지 분류와 객체 탐지 등 여러 분야에 활용됨에 따라, 자율주행자동차와 같이 안전필수시스템에 사용되는 CNN 모델의 성능은 신뢰할 수 있어야 한다. 이에 CNN 모델이 다양한 환경에서도 성능을 유지하는지 평가하기 위해 배경을 변경한 이미지를 생성하는 이미지 데이터 증강 도구를 개발한다. 이미지 데이터 증강 도구에 객체가 존재하는 이미지를 입력하면, 해당 이미지로부터 객체 이미지를 추출한 후 수집한 배경 이미지 내에 객체 이미지를 합성하여 새로운 이미지를 생성한다. CNN 모델 성능 평가 방법으로 개발한 도구를 사용하여 기존 테스트 이미지로부터 새로운 테스트 이미지를 생성하고, 생성한 새로운 테스트 이미지로 CNN 모델을 평가한다. 사례 연구로 Pascal VOC2007 테스트 데이터로부터 새로운 테스트 이미지를 생성하고, 새로운 테스트 이미지로 YOLOv3 모델을 평가하였다. 그 결과 기존 테스트 이미지의 mAP 보다 새로운 테스트 이미지의 mAP가 약 0.11 더 낮아지는 것을 확인하였다.

• 복식
• /
• 제61권9호
• /
• pp.1-14
• /
• 2011

The purpose of this study was to understand the relationship between one's social-self image and Online Avatar image. Influence of these virtual images on one's attitude toward real world and commitment to the virtual world was examined. In addition, the gender difference was examined. A structural equation model with social self image as exogenous variable and influence of Avatar as endogenous variable was designed. Real and ideal Avatar images were the mediating variable in the model. Survey questionnaire was developed and data from 425 respondents were analyzed. Results indicated that the conceptual model was a good fit to the data. Respondents who perceived their social self-images importantly were likely to have real images of Avatars. Ideal image and real image had significant on commitment to virtual world and attitudes toward the real world. For male respondents, social self image had stronger influence on real image of Avatar and ideal image had stronger influence on commitment to virtual world than female respondents.

• 한국의류학회지
• /
• 제44권3호
• /
• pp.556-571
• /
• 2020

This study predicts consumer preference for social clothing at work, excluding uniforms using the self-product congruence theory that also establishes a model to predict the preference for recommended products that match the consumer's own image. A total of 490 Korean male office workers participated in this study. Participants' self-image and the product images of 20 apparel items were measured using nine adjective semantic scales (namely elegant, stable, sincere, refined, intense, luxury, bold, conspicuous, and polite). A model was then constructed to predict the consumer preferences using a neural network with Python and TensorFlow. The resulting Predict Preference Model using Product Image (PPMPI) was trained using product image and the preference of each product. Current research confirms that product preference can be predicted by the self-image instead of by entering the product image. The prediction accuracy rate of the PPMPI was over 80%. We used 490 items of test data consisting of self-images to predict the consumer preferences for using the PPMPI. The test of the PPMPI showed that the prediction rate differed depending on product attributes. The prediction rate of work apparel with normative images was over 70% and higher than for other forms of apparel.

• International Journal of Computer Science & Network Security
• /
• 제21권3호
• /
• pp.83-93
• /
• 2021

COVID-19 poses a major risk to global health, highlighting the importance of faster and proper diagnosis. To handle the rise in the number of patients and eliminate redundant tests, healthcare information exchange and medical data are transmitted between healthcare centres. Medical data sharing helps speed up patient treatment; consequently, exchanging healthcare data is the requirement of the present era. Since healthcare professionals share data through the internet, security remains a critical challenge, which needs to be addressed. During the COVID-19 pandemic, computed tomography (CT) and X-ray images play a vital part in the diagnosis process, constituting information that needs to be shared among hospitals. Encryption and image steganography techniques can be employed to achieve secure data transmission of COVID-19 images. This study presents a new encryption with the image steganography model for secure data transmission (EIS-SDT) for COVID-19 diagnosis. The EIS-SDT model uses a multilevel discrete wavelet transform for image decomposition and Manta Ray Foraging Optimization algorithm for optimal pixel selection. The EIS-SDT method uses a double logistic chaotic map (DLCM) is employed for secret image encryption. The application of the DLCM-based encryption procedure provides an additional level of security to the image steganography technique. An extensive simulation results analysis ensures the effective performance of the EIS-SDT model and the results are investigated under several evaluation parameters. The outcome indicates that the EIS-SDT model has outperformed the existing methods considerably.