• The Journal of Korea Robotics Society
• /
• v.5 no.4
• /
• pp.339-348
• /
• 2010

One of the requirements for autonomous vehicles on off-road is to move stably in unstructured environments. Such capacity of autonomous vehicles is one of the most important abilities in consideration of mobility. So, many researchers use contact and/or non-contact methods to determine a terrain whether the vehicle can move on or not. In this paper we introduce an algorithm to classify terrains using visual information(one of the non-contacting methods). As a pre-processing, a contrast enhancement technique is introduced to improve classification of terrain. Also, for conducting classification algorithm, training images are grouped according to materials of the surface, and then Bayesian classification are applied to new images to determine membership to each group. In addition to the classification, we can build Traversability map specified by friction coefficients on which autonomous vehicles can decide to go or not. Experiments are made with Load-Cell to determine real friction coefficients of various terrains.

• International Journal of Internet, Broadcasting and Communication
• /
• v.10 no.1
• /
• pp.16-22
• /
• 2018

This paper presents a solution for personal profiling system based on user-oriented tracking. Here, we introduce a new way to identify and track humans by using two types of cameras: dome and face camera. Dome camera has a wide view angle so that it is suitable for tracking human movement in large area. However, it is difficult to identify a person only by using dome camera because it only sees the target from above. Thus, face camera is employed to obtain facial information for identifying a person. In addition, we also propose a new mechanism to locate human on targeted location by using grid-cell system. These result in a system which has the capability of maintaining human identity and tracking human activity (movement) effectively.

• Proceedings of the Korean Information Science Society Conference
• /
• 1998.10c
• /
• pp.435-437
• /
• 1998

영상에 대한 영역분할은 영상에 대한 인식 시스템에서 가장 중요하고도 어려운 분야로 알려져 있다. 주로 사용되는 방법은 화소중심기법과 영역중심기법이 사용되는데, 화소중심기법은 적은 시간이 걸리는데 비해 영역분할 효과가 떨어지고, 영역중심기법은 상대적으로 양질의 영역분할 효과를 얻을 수 있지만 많은 시간이 걸린다. 본 논문에서는 영역분할에 대한 방법으로 thresholding방법을 이용한 2단계로 이루어진 영역분할 방법을 제안한다. 제안된 방법은 화소의 전역정보와 지역정보를 모두 사용하여 기존의 전역 thresholding방법에 비해 향상된 영역 분할을 수행하고, 지역정보를 이용하는 영역중심 기법에 비해 시간을 단축하는 효과를 가지고 있다. 첫 번째 단계에서는 기존에 알려진 전역 thresholding방법을 사용하여 영역분할을 하고, 두 번째 단계에서는 영상에 대해 미리 알려진 사전지식을 이용하여 영역분할이 제대로 되지 않은 영역을 구분하여 해당 영역에 대해서만 thresholding작업을 수행한다. 사용된 영상은 자궁경부 세포진 영상으로 대상이 되는 영역은 자궁경부 세포의 핵으로 제한하였다.

• ETRI Journal
• /
• v.35 no.2
• /
• pp.207-217
• /
• 2013

This paper introduces the modular cellular neural network (CNN), which is a new CNN structure constructed from nine one-layer modules with intercellular interactions between different modules. The new network is suitable for implementing many image processing operations. Inputting an image into the modules results in nine outputs. The topographic characteristic of the cell interactions allows the outputs to introduce new properties for image processing tasks. The stability of the system is proven and the performance is evaluated in several image processing applications. Experiment results on texture segmentation show the power of the proposed structure. The performance of the structure in a real edge detection application using the Berkeley dataset BSDS300 is also evaluated.

• The Korean Journal of Soil Zoology
• /
• v.6 no.1_2
• /
• pp.1-6
• /
• 2001

Although the earthworm is an important animal species capable of regenerating missing body part, earthworm regeneration is not well understood at the tissue, cell and molecular levels. In order to understand the developmental changes of blastema during earthworm tail regeneration, the formation and development blastema during regeneration was investigated by histological analysis. Within 1 day after amputation, dediffentiating blastemal cells appeared in coelomic side of longitudinal muscle layer. At 3 days postamputation, proliferating blastemal cells migrated into coelum and blastema was formed beneath wound epithelium around 7 days after amputation. Segmentation of blastema was observed around 2 weeks after amputation followed by redifferentiation of nephridium, intestine, setae and septa.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2006.11a
• /
• pp.199-202
• /
• 2006

본 논문에서는 적응적 관심윈도우에 기반한 세포영상 세그먼테이션 기법을 제안한다. 명암지도를 이용하여 초기 관심윈도우를 생성하고, 초기 관심윈도우를 쿼드-트리 분할을 통해 실제 관심영역과 유사한 크기가 될 때까지 축소한다. 이렇게 생성된 적응적 관심윈도우는 세포영상에서 배경을 제거하고 관심영역 추출의 처리시간을 줄일 수 있다. 그리고 세그먼테이션과 관심영역의 분리를 위한 영역 병합 및 제거를 수행하여 최종적으로 정밀한 관심영역을 얻어낸다. 실험에서 제안된 기법은 세포영상의 관심영역을 효과적으로 분리하여 인간 시각과 유사한 향상된 세그먼테이션 결과를 보여준다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2010.11a
• /
• pp.697-700
• /
• 2010

피부 이미지의 여러 가지 특징들 중 주름은 피부의 상태를 판단하는 중요한 요소이다. 따라서 주름을 추적하기 위해 확대경으로 촬영된 원본 이미지에서 질감 대비 증가, 노이즈 제거 등의 전처리 과정을 수행한 후 Watershed 알고리즘을 이용하여 주름을 선분으로 표현하였다. 이렇게 생성된 주름의 깊이, 너비, 길이 등은 피부 분석 시 특징 정보로 이용할 수 있다. 또한 주름과 주름이 연결되어 이루는 다각형을 논문에서는 셀(Cell)이라고 정의하는데 그것의 크기나 개수 같은 정보도 추출할 수 있게 된다. 그러나 주름으로 만들어진 셀들은 실제와 다르게 과분할 되는 경향을 보인다. 과분할 된 셀들은 잘못된 정보를 제공하기 때문에 피부 상태를 판단하는 결과의 정확도를 떨어뜨린다. 본 논문에서는 이러한 문제점을 인지하고 차후 정확한 셀 정보를 획득하기 위한 확장성 측면에서 각 셀들을 개체화시키고 과분할 된 셀을 검출하는 방법을 제안한다.

• Journal of Advanced Information Technology and Convergence
• /
• v.9 no.1
• /
• pp.89-102
• /
• 2019

Collagen can be used in building artificial skin replacements for treatment of burns and towards the reconstruction of bone as well as researching cell behavior and cellular interaction. The strength of collagen in connective tissue rests on the characteristics of collagen fibers. 3D confocal imaging of collagen fibers enables the characterization of their spatial distribution as related to their function. However, the image stacks acquired with confocal laser-scanning microscope does not clearly show the collagen architecture in 3D. Therefore, we developed a new method to reconstruct, visualize and characterize collagen fibers from fluorescence confocal images. First, we exploit the tensor voting framework to extract sparse reliable information about collagen structure in a 3D image and therefore denoise and filter the acquired image stack. We then propose to segment the collagen fibers by defining an energy term based on the Hessian matrix. This energy term is minimized by a min cut-max flow algorithm that allows adaptive regularization. We demonstrate the efficacy of our methods by visualizing reconstructed collagen from specific 3D image stack.

• Proceedings of the Korean Society of Embryo Transfer Conference
• /
• 2002.11a
• /
• pp.118-118
• /
• 2002

Programmed cell death (PCD) is thought as a well-controlled process by which unwanted cells are selectively eliminated. During the last decade many researches have elucidated molecules and their interactions involved in cell death by using largely in vitro induction of cell death or survival signals in a more defined manner, While these critical information and novel findings provide us with clearer understanding of mechanisms underlying cell death, it does by no means explain how PCD occurs and which cells or tissues are affected during normal embryonic development in vivo. In this study, we used zebrafish to examine whether the PCD is occurring selectively or randomly in developing embryos by whole mount in situ TUNEL analysis with specific markers for neural cells. The result revealed that the degree and distribution of TUNEL staining varied considerably throughout gastrulation stage, and there was also a number of TUNEL-negative embryos. Most of TUNEL-positive cells were scattered randomly throughout the blastoderm. During the gastrulation stage about 75 % of the embryos analyzed exhibited more than 5 TUNEL-positive cells. As the dorsal epiblast begins to thicken rather abruptly near the end of gastrulation, TUNEL-positive cells were mainly located along the dorsal side. Although there were some variations in TUNEL staining during segmentation and pharyngeal stages, TUNEL staining continued to be localized to the central nervous system, and was also detected in the sensory organs, trigeminal ganglions, and the primary sensory neurons. High levels of the cell death in developing brain between 20-somite and prim-6 stages are thought to play a role in the morphogenesis and organization of the brain. At prim-16 stage, cell death is considerably reduced in the brain region. Dying cells are mainly localized to the prospective brain region where ectodermal cells are about to initiate neurogenesis. As development progressed, high levels and more reproducible patterns of cell death were observed in the developing nervous system. Intensive TUNEL staining was restricted to the trigeminal ganglions, the primary sensory neurons, and sensory organs, such as olfactory pits and otic vesicles. Thus, PCD patterning in zebrafish embryos occurs randomly at early stages and becomes restricted to certain region of the embryos. The spatio-temporal pattern of PCD during the early embryonic development in zebrafish will provide basic information for further studies to elucidate genes involved in. regulation of PCD largely unknown in vivo during vertebrate embryogenesis.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.18 no.3
• /
• pp.385-396
• /
• 1993

In this paper, a improved method of digital image analysis required in basic medical science for diagnosis of cells was proposed. The object image was the thyroid gland cell image, and the purpose was automatic discrimination of three classes cells(normal cell, follicular neoplastic cells, and papillary neoplastic cells) by difference of chromatin patterns. To segment the cell nucleus from background, the region segmentation algorithm by edge tracing was proposed. And feature parameter was obtained from discrete Fourier transformation of image. After construct a feature sample group of each cells, experiment of discrimination was executed with any verification cells. As a consequency of using features proposed in this paper, get a better recognition rate(70-90%) than previously reported papers, and this method give shape to get objectivity and fixed quantity in diagnosis of cells, The methods described in this paper be used immediately for discrimination of neoplastic cells.