• Title/Summary/Keyword: deep color

Search Result 566, Processing Time 0.024 seconds

Game Sprite Generator Using a Multi Discriminator GAN

  • Hong, Seungjin;Kim, Sookyun;Kang, Shinjin
    • KSII Transactions on Internet and Information Systems (TIIS)
    • /
    • v.13 no.8
    • /
    • pp.4255-4269
    • /
    • 2019
  • This paper proposes an image generation method using a Multi Discriminator Generative Adversarial Net (MDGAN) as a next generation 2D game sprite creation technique. The proposed GAN is an Autoencoder-based model that receives three areas of information-color, shape, and animation, and combines them into new images. This model consists of two encoders that extract color and shape from each image, and a decoder that takes all the values of each encoder and generates an animated image. We also suggest an image processing technique during the learning process to remove the noise of the generated images. The resulting images show that 2D sprites in games can be generated by independently learning the three image attributes of shape, color, and animation. The proposed system can increase the productivity of massive 2D image modification work during the game development process. The experimental results demonstrate that our MDGAN can be used for 2D image sprite generation and modification work with little manual cost.

Deep Learning based Color Restoration of Corrupted Black and White Facial Photos (딥러닝 기반 손상된 흑백 얼굴 사진 컬러 복원)

  • Woo, Shin Jae;Kim, Jong-Hyun;Lee, Jung;Song, Chang-Germ;Kim, Sun-Jeong
    • Journal of the Korea Computer Graphics Society
    • /
    • v.24 no.2
    • /
    • pp.1-9
    • /
    • 2018
  • In this paper, we propose a method to restore corrupted black and white facial images to color. Previous studies have shown that when coloring damaged black and white photographs, such as old ID photographs, the area around the damaged area is often incorrectly colored. To solve this problem, this paper proposes a method of restoring the damaged area of input photo first and then performing colorization based on the result. The proposed method consists of two steps: BEGAN (Boundary Equivalent Generative Adversarial Networks) model based restoration and CNN (Convolutional Neural Network) based coloring. Our method uses the BEGAN model, which enables a clearer and higher resolution image restoration than the existing methods using the DCGAN (Deep Convolutional Generative Adversarial Networks) model for image restoration, and performs colorization based on the restored black and white image. Finally, we confirmed that the experimental results of various types of facial images and masks can show realistic color restoration results in many cases compared with the previous studies.

Detection and Classification of Leaf Diseases for Phenomics System (피노믹스 시스템을 위한 식물 잎의 질병 검출 및 분류)

  • Gwan Ik, Park;Kyu Dong, Sim;Min Su, Kyeon;Sang Hwa, Lee;Jeong Hyun, Baek;Jong-Il, Park
    • Journal of Broadcast Engineering
    • /
    • v.27 no.6
    • /
    • pp.923-935
    • /
    • 2022
  • This paper deals with detection and classification of leaf diseases for phenomics systems. As the smart farm systems of plants are increased, It is important to determine quickly the abnormal growth of plants without supervisors. This paper considers the color distribution and shape information of leaf diseases, and designs two deep leaning networks in training the leaf diseases. In the first step, color distribution of input image is analyzed for possible diseases. In the second step, the image is first partitioned into small segments using mean shift clustering, and the color information of each segment is inspected by the proposed Color Network. When a segment is determined as disease, the shape parameters of the segment are extracted and inspected by proposed Shape Network to classify the leaf disease types in the third step. According to the experiments with two types of diseases (frogeye/rust and tipburn) for apple leaves and iceberg, the leaf diseases are detected with 92.3% recall for a segment and with 99.3% recall for an input image where there are usually more than two disease segments. The proposed method is useful for detecting leaf diseases quickly in the smart farm environment, and is extendible to various types of new plants and leaf diseases without additional learning.

A Study on the Sketch of Trikaya Banner Painting in the Suta-sa Temple (수타사 삼신불괘불도(三身佛掛佛圖) 초본(草本) 연구)

  • Kim, Chang Kyun
    • Korean Journal of Heritage: History & Science
    • /
    • v.42 no.4
    • /
    • pp.112-131
    • /
    • 2009
  • The Trikaya Banner Painting in the Suta-sa Temple at Hongcheon-gun, Gangwon-do draws attention as it was painted not on flax but on paper, and used the water color painting technique on the sketch rather than the deep color painting technique, which is most common in Buddhist paintings created during the Chosun Dynasty. Nevertheless, there is not any information on the creation of the Trikaya Banner Painting in the painting record on the painting, in Sutasasajeok(壽陀寺史蹟), or in Sutasagogirok(壽陀寺古記錄), so it is uncertain when the painting was created. Furthermore, because it was not drawn by the deep color painting technique, it has been difficult to compare it with other banner paintings. For these reasons, the Trikaya Banner Painting has been studied little except brief introduction. In recent preservation treatment that removed multiple-layered paper from the back of the painting, however, an inked inscription written on Korean paper 118cm high and 87.5cm wide was discovered on the back. It is a kind of placard notifying a number of acts prohibited in order to follow Buddha's teachings correctly, and was found to have been written on April 15, 1690. The inked inscription is a very valuable material for estimating the creation date of the Suta-sa Trikaya Banner Painting, and provides crucial clues for approaching the contents and nature of the painting more precisely. When the image, form, and style of the Suta-sa Trikaya Banner Painting were examined and its creation date was estimated based on the inked inscription, first, the painting is presumed to have been created in around 1690 as suggested by 'the placard' attached on the back instead of a painting record. Second, the painting is highly likely to be the first standing Trikaya banner painting showing the basic icons of Trikaya banner paintings in the Chosun Dynasty since the Trikaya Banner Painting in the Gap-sa Temple in Gongju (1650). Furthermore, considering the shape of the Trikaya in the painting, screen composition, background treatment, solemn and affectionate facial expression, harmonious and adequate body proportion, etc., the painting is believed to have had a considerable influence not only on Trikaya banner paintings of similar style in the 18thcentury but also on deep-color Trikaya banner paintings in the 19thcentury. Third, although the Suta-sa Trikaya Banner Painting is not acompleted work but a sketch, it exhibits the typical water color painting technique in which the strokes are clearly visible. Thus, it is considered highly valuable in understanding and analyzing stroke styles and in studying the history of Buddhist paintings. As there are not many extant banner paintings of the same style in form and expression technique as the Suta-sa Temple Trikaya Banner Painting, this study could not make thorough comparative analysis of the work, but still it is meaningful in that it laid the ground for research on standing Trikaya banner paintings in the 18thand 19thcenturies in the Chosun Dynasty.

Comparison of Quality Characteristics of Salted Muskmelon with Deep Seawater Salt, Sun-dried and Purified Salts (해양심층수염, 천일염 및 정제염을 이용한 참외절임시 품질특성 비교)

  • 이기동;김숙경;김정옥;김미림
    • Journal of the Korean Society of Food Science and Nutrition
    • /
    • v.32 no.6
    • /
    • pp.840-846
    • /
    • 2003
  • Physiochemical and sensory properties of muskmelon pickles were investigated under salting condition using deep seawater salt, sun-dried salt and purified salt. The changes of weight, shrinkage rate salinity and pH of muskmelon during salting time were almost nothing shown. Turbidity of pickle juice was the highest in salt solution using sun-dried salt and was almost nothing shown in deep seawater salt and purified salt. Solid content of pickle juice was higher in pickle juice using sun-dried salt than deep seawater salt and purified salt. The calcium content of muskmelon salted by deep seawater salt and sun-dried salt were increased to 4.3 times and 3.7 times, respectively. Hardness became a little higher on muskmelon using deed seawater salt ann sun dried salt than purified salt. In organoleptic result, color, flavor, taste, texture and overall palatability of muskmelon salted by deep seawater salt became higher in other salt solution (sun-friend salt and purified salt).

Preparation of Novel Fused Ring Spiro[benzotetraphene-fluorene] Derivatives and Application for Deep-Blue Host Materials

  • Kim, Min-Ji;Lee, Chil-Won;Gong, Myoung-Seon
    • Bulletin of the Korean Chemical Society
    • /
    • v.35 no.6
    • /
    • pp.1639-1646
    • /
    • 2014
  • A series of novel fused-ring spiro compounds, spiro[benzo[ij]tetraphene-7,9'-fluorene] (SBTF) derivatives containing an end-capping aryl substituent at both the C3 and C10-positions hasbeen designed and synthesized via multi-step Suzuki coupling reactions. 3-(1-Naphthyl)-10-phenylSBTF (1N-PSBTF), 3-(2-naphthyl)-10-phenylSBTF (2N-PSBTF) and 3-[4-(1-naphthyl)phenyl]-10-phenylSBTF (NP-PSBTF) showed improved glass transition temperatures ($T_g$) with good thermal stability. Their photophysical, electrochemical, and electroluminescent properties were investigated and were used to construct blue organic light emission diodes (OLEDs). The typical OLED devices showed excellent performance; the NP-PSBTF-based device exhibited highly efficient deep blue-light emission with a maximum efficiency of 5.27 cd/A (EQE, 4.63%) with CIE (x = 0.133, y = 0.144). According to these characteristics, these deep-blue light emitting materials have sufficient potential for fluorescent OLED applications.

Low-light Image Enhancement Method Using Decomposition-based Deep-Learning (분해 심층 학습을 이용한 저조도 영상 개선 방식)

  • Oh, Jong-Geun;Hong, Min-Cheol
    • Journal of IKEEE
    • /
    • v.25 no.1
    • /
    • pp.139-147
    • /
    • 2021
  • This paper introduces an image decomposition-based deep learning method and loss function to improve low-light images. In order to remove color distortion and halo artifact, illuminance channel of an input image is decomposed into reflectance and luminance channels, and a decomposition-based multiple structural deep learning process is applied to each channel. In addition, a mixed norm-based loss function is described to increase the stability and remove blurring in reconstructed image. Experimental results show that the proposed method effectively improve various low-light images.

A Research of Ink and Wash Elements on the 3D Animation Film <Deep Sea>

  • Biying Guo;Xinyi Shan;Jeanhun Chung
    • International Journal of Internet, Broadcasting and Communication
    • /
    • v.15 no.3
    • /
    • pp.82-87
    • /
    • 2023
  • <Deep Sea> is an 3D animated film that stands out for its exceptional special effects and distinctive artistic style. The film employs a multitude of dazzling and vibrant ink particles, creating a strong sense of three-dimensionality and weightlessness, while simultaneously portraying a dreamlike and elegant representation of a deep sea ink painting. Furthermore, through the utilization of fragmented stream of consciousness narrative technique, the film establishes a unique artistic effect infused with a Chinese atmosphere. This paper by analyzing the unique particle ink art style and color and stream of consciousness narrative methods in film, this paper discusses the innovative art style generated by traditional ink art style combined with three-dimensional technology, and the integration of traditional ink art ideas and artistic conception in animated films. The objective is to cultivate a new ink art style and prove the importance of traditional cultural expression in animated films, while providing new perspectives for the future application of traditional art in animation.

Analysis on Color Shaping the Space of Animated Film Coco (애니메이션영화 <코코> 색상이 공간의 모델링 분석)

  • Yue, Xiao Ling
    • Journal of Korea Entertainment Industry Association
    • /
    • v.13 no.3
    • /
    • pp.343-350
    • /
    • 2019
  • As an important visual element, color in animated films can not only depicts the reality, but also conveys emotion and meanings, which play an important role in shaping the space of animated films. As one of the most outstanding animated films, Coco uses bold and unique color language to shape the film space. It uses the national color to blend the Mexican culture into the story space making it full of Mexican regional and cultural connotation, uses color contrast to shape the physical space, highlighting the three-dimensional and deep sense of the physical space. And also uses color association to create the psychological space, leading audience into the spiritual world of the film. Based on the analysis of the Coco, we an see that when shape the film space of animation, the film maker should give full play to the characteristics and function of color, and apply different color language to shape different space according to the role of the film space, so as to expand and deepen the connotation and expressiveness of film space.

Development of Late Bolting and New Deep Red Leaf with Wrinkled Lettuce 'Chunpungjeokchukmyeon' (추대가 늦고 색깔이 진한 새로운 잎상추 '춘풍적축면' 육성)

  • Jang, Suk-Woo;Hur, Youn-Young;Choi, Mi-Ja;Kwon, Young-Seok;Kim, Jeom-Sun;Lee, Jong-Nam;Lee, Eung-Ho;Seo, Myeong-Hun;Park, Jae-Ho;Jang, Ik;Jang, Mi-Hyang;Hwang, Hae-June;Ko, Sun-Bo
    • Korean Journal of Breeding Science
    • /
    • v.42 no.6
    • /
    • pp.627-631
    • /
    • 2010
  • A new cultivar of lettuce (Lactuca sativa L.) with wrinkled traverse elliptic and deep red leaf, 'Chunpungjeokchukmyeon' which has late bolting and deep red expression leaf color was developed from a cross between 'Pojabijeokchukmyeon' (red leaf color and late bolting) and 'Meokchima' (Deep red and low yield). The cross and selection for advanced lines had been done by the pedigree method during 2000-2007. The advanced lines were evaluated for yield and adaptability at several locations in Korea (Gangwon-do, Gyeonggi-do, Chungcheongbuk-do, Jeollabuk-do, Gyeongnam-do, and Jeju-do) from 2008 to 2009. The 'Chunpungjeokchukmyeon' has gray seed color and traverse elliptic leaves. The type of matured stage is medium shape between 'chukmyeon' and 'chima' leaf lettuce. Compared to 'Dukseomjeokchukmyeon', marketable yield of 'Chunpungjeokchukmyeon' was higher by 6% (at 372 g per plant) and 'Chunpungjeokchukmyeon' has particularly improved expression of deep red leaf color in high temperature cultivation in the field. The shelf-life of 'Chunpungjeokchukmyeon' was three weeks longer than 'Dukseomjeokchukmyeon' at 4$^{\circ}C$. The anthocyanin content of 'Chunpungjeokchukmyeon' was higher than that of 'Dukseomjeokchukmyeon' with 17.5 mg/100g. The BSL (latucin+8-deoxylactucin+lactucopicrin) content of 'Chunpungjeokchukmyeon' is lower than that of 'Dukseomjeokchukmyeon'. Furthermore, its taste is better, more crispy, and sweeter than those of 'Dukseomjeokchukmyeon'. So we recommend that new cultivar 'Chunpungjeokchukmyeon' can be suitable for cultivation in spring season than summer season.