• 대한임베디드공학회논문지
• /
• 제12권2호
• /
• pp.97-103
• /
• 2017

Thermal imaging is mainly used in military equipment required for night observation. In particular, technologies of uncooled thermal imaging detectors are being developed as applied to low-cost night observation system. Many system integrators require different specifications of the uncooled thermal imaging camera but their development time is short. In this approach, EOSYSTEM has developed a small size, TEC-less uncooled thermal imaging camera module with $32{\times}32mm$ size and low power consumption. Both domestic detector and import detector are applied to the EOSYSTEM's thermal imaging camera module. The camera module contains efficient infrared image processing algorithms including : Temperature compensation non-uniformity correction, Bad/Dead pixel replacement, Column noise removal, Contrast/Edge enhancement algorithms providing stable and low residual non-uniformity infrared image.

• 한국산업융합학회 논문집
• /
• 제23권2_2호
• /
• pp.255-261
• /
• 2020

We consider the problem of autonomously flying a quadcopter in indoor environments. Navigation in indoor settings poses two major issues. First, real time recognition of the marker captured by the camera. Second, The combination of the distributed images is used to determine the position and orientation of the quadcopter in an indoor environment. We autonomously fly a miniature RC quadcopter in small known environments using an on-board camera as the only sensor. We use an algorithm that combines data-driven image classification with image-combine techniques on the images captured by the camera to achieve real 3D localization and navigation.

• 전자공학회논문지B
• /
• 제32B권12호
• /
• pp.1581-1590
• /
• 1995

Recently, advanced video coding techniques using segmentation technique have been actively researched as candidates for video coding of MPEG-4 standard. The conventional segmentation techniques are unsuitable for real-time process because they have sequential structure. In this paper, we propose a new image segmentation technique using competitive learning neural network for vector quantization. The proposed segmentation procedure consist of prefiltering, primary and secondary segmentation, and a small region ellimination process. Primary segmentation segments input image in detail. Secondary segmentation merges similar region using a repetitive FSCL(Frequency sensitive competive learning) neural network. In this process, it is possible to segment an image from high resolution to low resolution by adjusting the number of repetition. Finally, small regions are merged into adjacent regions. Experimental results show that the procedure described yields reconstructed images of reasonably acceptable quality at bit rates of 0. 25 - 0.3 bit/pel.

• 한국멀티미디어학회논문지
• /
• 제24권11호
• /
• pp.1508-1517
• /
• 2021

In CCTV environments, a lot of learning image data is required to monitor illegal dumping of garbage with a typical image-based object detection using deep learning method. In this paper, we propose a system to monitor unauthorized dumping of garbage by learning the articular points of the person using only a small number of images without immediate use of the image for deep learning. In experiment, the proposed system showed 74.97% of garbage dumping detection performance with only a relatively small amount of image data in CCTV environments.

• 한국해양공학회지
• /
• 제16권6호
• /
• pp.32-36
• /
• 2002

This paper presents the results of wave profile detection from video image using the Mexican hat function. The Mexican hat function has been extensively used in the field of signal processing to detect discontinuity in the images. The analysis was done on the numerical image and video images of waves that were taken in the small wave flume. The results show that the Mexican hat function is an excellent tool for wave profile detection.

• 대한기계학회논문집B
• /
• 제24권6호
• /
• pp.760-768
• /
• 2000

Filter efficiency of electrically charged particle in uncharged fibrous filter was measured. In previous studies, the effect of charged particle on filter efficiency was investigated but there was difficulty in measuring of image effect that is appeared at the charged small particle. We could easily measure the image effect with charging small particles by photoelectric charging. The spark discharge aerosol generator and a differential mobility analyzer (DMA) were used to generate sub-micron monodisperse particles (${\leq}200$ nm). The generated particles were charged in photoelectric charging process using ultraviolet lamp and electric field. The filter efficiency of the charged particles, classified by another DMA, was measured in filter tester using a condensation nucleus counter (CNC) as function of particle diameter, particle charge and airflow velocity. It is shown that the filter efficiency increases with increasing charge number of the particle and is affected by particle size and flow velocity. Single fiber filter efficiency mainly depends on image force parameter and peclet number. The peclet number was not considered at previous other papers. We propose a modi fied experimental correlation as function of image force parameter and peclet number.

• 한국정밀공학회:학술대회논문집
• /
• 한국정밀공학회 2002년도 추계학술대회 논문집
• /
• pp.404-406
• /
• 2002

In this paper, the method of image feature extraction is proposed. This method employ the energy field analysis, outlier removal algorithm and ring projection. Using this algorithm, we achieve rotation-translation-scale invariant feature extraction. The force field are exploited to automatically locate the extrema of a small number of potential energy wells and associated potential channels. The image feature is acquired from relationship of local extrema using the ring projection method.

• 융합신호처리학회논문지
• /
• 제12권2호
• /
• pp.76-82
• /
• 2011

본 논문은 다중 해상도 영상에서 3차 페이싯 모델을 이용하여 적외선 영상의 원거리에 위치하고 있는 초소형 표적의 위치와 크기를 검출하기 위한 방법을 제안한다. 먼저, 원 영상을 점차 축소하여 여려 단계의 다중 해상도의 영상들로 구성한다. 각 단계에서의 다중 해상도 영상들에 대해 페이싯 모델과 국부 극대 조건을 적용하여 초소형 표적의 위치를 검출한다. 다중 해상도 영상에서 각 페이싯 모델의 국부 극대값을 의미하는 $D_2$값 중 최대 크기를 가지는 위치를 표적의 위치라고 평가한다. 이 경우 각 단계의 다중 해상도 영상들에 대해 크기가 다른 표적의 검출이 가능하게 된다. 본 논문에서 제안한 초소형 표적 검출 방법은 초소형 표적이 있는 다양한 적외선 영상에서 실험하였다. 기존의 페이싯 모델을 이용한 방법에서는 하나의 마스크만 적용시킨 것에 반해 제안된 방법은 하나의 마스크를 다중 해상도 영상에서 적용하였다. 고정된 마스크를 다중 해상도 영상에 적용함으로써 마스크의 크기를 달리하는 효과를 확인하였고 그에 따라 검출하는 표적의 크기도 다름을 확인하였다. 이를 이용해서 표적의 위치뿐만 아니라 크기도 검출할 수 있음을 확인하였다.

• 인터넷정보학회논문지
• /
• 제6권2호
• /
• pp.59-68
• /
• 2005

본 논문에서는 3차원 영상의 압축 전송을 위하여 3차원 영상에 대해 3차원 DCT를 수행하고 원 영상과의 비교에 따라 영상의 3차원 블록들을 계층적으로 분할하여 각 블록에 대해 유한상태 벡터 양자화를 수행하는 알고리듬을 제안한다. 3차원 DCT의 계수 특징을 이용하여 영상을 크기가 큰 배경 블록과 크기가 작은 윤곽선 블록으로 계층적으로 분할하고, 블록 계층분할 정보를 전송한다. 각 계층에 속한 블록들에 대해 따로 부호책을 설계하고 부호 비트 수를 줄이기 위해 유한상태 벡터양자화를 이용하여 부호단어의 인덱스를 계층 분할 정보와 함께 전송한다. Small Lobster와 Head 영상에 대하여 본 알고리듬으로 부호화했을 때 기존의 HFSVQ를 이용한 알고리듬보다는 각각 1.91 dB과 1.47 dB만큼 더 좋은 영상의 화질을 얻을 수 있었다.

• 한국군사과학기술학회지
• /
• 제19권3호
• /
• pp.288-293
• /
• 2016

Electronic Image Stabilization(EIS) is widely used as a technique for correcting a shake of an image. The case requiring the EIS function has been increased in high magnification thermal image observation on portable military equipment. Projection Algorithm(PA) for EIS is easy to implement but its performance is sensitive to the projection area. Especially, projection profiles of thermal image have very modest change and are difficult to extract image shifts between frames. In this paper, we proposed algorithm to extract a feature image for the thermal image and compared Block Matching Algorithm(BMA) with PA using our proposed feature image. When using our proposed feature image, BMA was simply implemented using FPGA's internal small memory. And we were able to obtain 30 % PSNR improved results compared to PA.