• Journal of Broadcast Engineering
• /
• v.27 no.1
• /
• pp.31-43
• /
• 2022

Traditional video compression has developed so far based on hybrid compression methods through motion prediction, residual coding, and quantization. With the rapid development of technology through artificial neural networks in recent years, research on image compression and video compression based on artificial neural networks is also progressing rapidly, showing competitiveness compared to the performance of traditional video compression codecs. In this paper, a new method capable of improving the performance of such an artificial neural network-based video compression model is presented. Basically, we take the rate-distortion optimization method using the auto-encoder and entropy model adopted by the existing learned video compression model and shifts some components of the latent information that are difficult for entropy model to estimate when transmitting compressed latent representation to the decoder side from the encoder side, and finally compensates the distortion of lost information. In this way, the existing neural network based video compression framework, MFVC (Motion Free Video Compression) is improved and the BDBR (Bjøntegaard Delta-Rate) calculated based on H.264 is nearly twice the amount of bits (-27%) of MFVC (-14%). The proposed method has the advantage of being widely applicable to neural network based image or video compression technologies, not only to MFVC, but also to models using latent information and entropy model.

• Journal of the Korean Society for Precision Engineering
• /
• v.21 no.12
• /
• pp.84-91
• /
• 2004

A CCD camera with a laser stripe was applied to realize the automatic weld seam tracking in GMAW. It takes relatively long time to process image on-line control using the basic Hough transformation, but it has a tendency of robustness over the noises such as spatter and arc tight. For this reason, it was complemented with adaptive Hough transformation to have an on-line processing ability for scanning specific weld points. The adaptive Hough transformation was used to extract laser stripes and to obtain specific weld points. The 3-dimensional information obtained from the vision system made it possible to generate the weld torch path and to obtain the information such as width and depth of weld line. In this study, a neural network based on the generalized delta rule algorithm was adapted for the process control of GMA, such as welding speed, arc voltage and wire feeding speed.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.32B no.1
• /
• pp.120-126
• /
• 1995

The diffusion neural network can be cfficiently applied to the Gaussian processing. For example, a difference of two Gaussians(DOG) is performed by this network with ease. In this paper, we model a neural network to perform the function /t(.del.${\Delta}^{2}$G) by using the diffusion neural network. This model is used to detect the edges of moving target in image. By this model not only moving target is separated from stationary background but also their trajectories are obtained using accumulated past information in the diffusion neural network. Furthermore this model needs a small number of connections per cell and the connection weights are fixed-valued. Therefore its hardware can be easily implemented with simple structure.

• The Journal of the Acoustical Society of Korea
• /
• v.13 no.1
• /
• pp.42-48
• /
• 1994

Recently pictorial information processing has become increasingly important So, this paper described that feature points of fingerprint used to recognize fingerprints for identification in a government or arresting criminals in an institution like a police station related to crime. We apply an algorithm based on minimization of fuzzy theory to thinning and then the image into a certain size of squares. We have got some information about cores and deltas from the data encoding Into one of the eight directional codes. We could extract about $80\%$ feature points as the result of the experiment.

• Journal of the Korean Graphic Arts Communication Society
• /
• v.26 no.2
• /
• pp.65-77
• /
• 2008

The color reproduction of digital still camera does not, in general, match those of the final prints. Because color gamut of these devices is different, it is therefore necessary to take account of a way to match. The way uses the optimized profile to print an image. This paper proposed a way to create the input profile of digital still camera for standardization printing process. The results of proposed way showed that for input profiles equivalent, good results relatively. In this paper, an experiment was done where the illumination sources used as the standard illumination 5200K and illuminated at a $45^{\circ}$ angle in the best illumination efficiently. The white balance was in mode 'custom': aperture F11, exposure time 1/60s, ISO50, focal length 80mm. The images were exported and saved as 16bit RGB tiff(AdobeRGB, sRGB, ProphotoRGB) images. To do the test, the RGB values of the RGB tiff images are processed through the ICC input profile to arrive at processed $CIEL^*a^*b^*$ values. A profiling tool such as ProfileMaker 5.0 and Monacoprofile 4.8 are used to do this. The processed CIEL*a*b* values are compared to the reference CIEL*a*b* values and these two values are used to calculate a ${\Delta}E$.

• Korean Journal of Optics and Photonics
• /
• v.20 no.1
• /
• pp.53-56
• /
• 2009

We applied an ellipsometric technique to get quantitative information about the thickness and refractive index of human Mesenchymal Stem Cells (hMSCs). The images of ellipsometric constants $\Delta$, $\Psi$ for the nucleus region and for the cell body region of hMSCs were obtained by using an Imaging Ellipsomter (IE) for their in vitro state. A numerical inversion method was applied to deduce the refractive index and the thickness of hMSCs from the measured $\Delta$, $\Psi$. Thus the images of the refractive index and those of the thickness of hMSCs for the nucleus region and for the cell body region are reported.

• Journal of the Korea Furniture Society
• /
• v.10 no.1
• /
• pp.1-12
• /
• 1999

Of the vast number of plant taxa in the world, the wood is one of the most useful resources. It is important to identify the fibers of wood and pulp for the plant taxonomy and for the uses, but we do not have enough information on them, on them, especially for the computerizd data. The fiber identification is one of the difficult tasks. In addition to the plant taxonomy and the fiber-using industries, such identification is also important in many other fields, including education. document examiners, etc. For these purpose, the fibers should be exactly distinguished. The TISS system I have programed to identify various woods would also be useful in the identification of fibers by the genus and species in the features of unknown samples and in searching the features of a species based on its scientific name. Such searching programs are being developed in many other countries with a view to searching for the species name by using the features of the cells of the woody materials. With the survey of all the available literature, the features of the fibers of 124 species both of softwood and hardwood were examined under the electron and optical microscopies. Each species were coded and carded by the feature, and the databases were built. The microscopic were inputted into a personal computer program called and by a slide film scanner. The new computer program called TISS 2 was developed using C computer language. Korean language fonts were added to the TISS 2. The TISS 2 can be in adding and searching a image of fiber features both of a known fiber and an unknown fiber. The databases were corded for the DELTA system with was developed by Dallwitz and Paine in Australia, 1986.

• Journal of the Korean Graphic Arts Communication Society
• /
• v.27 no.1
• /
• pp.1-14
• /
• 2009

The color reproduction of digital still camera does not, in general, match those of the final output device. Because color gamut of these devices is different, it is therefore necessary to take account of a way to match. The way uses the optimized profile to output device an image. This paper proposed a way to create the input profile of digital still camera for standardization soft proofing process. The results of proposed way showed that for input profiles equivalent, good results relatively. In this paper, an experiment was done where the illumination sources used as the standard illumination 5200K and illuminated at a $45^{\circ}$ angle in the best illumination efficiently. The white balance was in mode 'custom' : aperture F11, exposure time 1/60s, ISO50, focal length 80mm. The images were exported and saved as 16bit RGB TIFF(AdobeRGB, sRGB, ProphotoRGB) images. To do the test, the RGB values of the RGB TIFF images are processed through the ICC input profile to arrive at processed $CIEL^*a^*b^*$ values. A profiling tool such as ProfileMaker 5.0 and Monacoprofile 4.8 are used to do this. The processed $CIEL^*a^*b^*$ values are compared to the reference $CIEL^*a^*b^*$ values and these two values are used to calculate a ${\Delta}E{^*}_{ab}$.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• v.9 no.2
• /
• pp.393-398
• /
• 2005

본 논문에서는 ART2 기반 자가 생성 지도 학습 알고리즘을 이용한 운송 컨테이너 식별자 인식 시스템을 제안한다. 일반적으로 운송 컨테이너의 식별자들은 글자의 색이 검정색 또는 흰색으로 이루어져 있는 특징이 있다. 이러한 특성을 고려하여 원 컨테이너 영상에 대해 검은색과 흰색을 제외한 모든 부분을 잡음으로 처리하기 위해 퍼지를 이용한 잡은 판단 방법을 적용하여 식별자 영역과 잡음을 구별한다. 식별자 영역을 제외한 잡음 영역을 전체 영상의 평균 픽셀값으로 대체시킨다. 그리고 Sobel 마스크를 이용하여 에지를 검출하고, 추출된 에지를 이용하여 수직 블록과 수평 블록을 검출하여 컨테이너의 식별자 영역을 추출하고 이진화한다. 이진화된 식별자 영역에 대해 검정색의 빈도수를 이용하여 흰바탕과 민바탕을 구분하고 8방향 윤곽선 추적 알고리즘을 적용하여 개별 식별자를 추출한다. 개별 식별자 인식을 위해 ART2 기반 자가 생성 지도 학습 알고리즘은 입력층과 은닉층 사이에 ART2를 적용하여 은닉층의 노드를 생성하고, 은닉층과 출력층 사이에 일반화된 델타 학습 방법과 Delta-bar-Delta 알고리즘을 적용하여 학습 성능을 개선한다. 실제 컨테이너 영상을 대상으로 실험한 결과, 기존의 식별자 추출 방법보다 제안된 식별자 추출 방법이 개선되었다. 그리고 기존의 식별자 인식 알고리즘보다 제안된 ART2 기반 자가 생성 지도 학습 알고리즘이 식별자의 학습 및 인식에 있어서 우수한 성능이 있음을 확인하였다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.18 no.1
• /
• pp.15-29
• /
• 2024

A fetoscope is an optical endoscope, which is often applied in fetoscopic laser photocoagulation to treat twin-to-twin transfusion syndrome. In an operation, the clinician needs to observe the abnormal placental vessels through the endoscope, so as to guide the operation. However, low-quality imaging and narrow field of view of the fetoscope increase the difficulty of the operation. Introducing an accurate placental vessel segmentation of fetoscopic images can assist the fetoscopic laser photocoagulation and help identify the abnormal vessels. This study proposes a method to solve the above problems. A novel encoder-decoder network with a dual-path structure is proposed to segment the placental vessels in fetoscopic images. In particular, we introduce a channel attention mechanism and a continuous convolution structure to obtain multi-scale features with their weights. Moreover, a switching connection is inserted between the corresponding blocks of the two paths to strengthen their relationship. According to the results of a set of blood vessel segmentation experiments conducted on a public fetoscopic image dataset, our method has achieved higher scores than the current mainstream segmentation methods, raising the dice similarity coefficient, intersection over union, and pixel accuracy by 5.80%, 8.39% and 0.62%, respectively.