• 로봇학회논문지
• /
• 제18권3호
• /
• pp.271-280
• /
• 2023

FMCW (Frequency Modulated Continuous Wave) radar system is widely used in autonomous driving and navigation applications due to its high detection capabilities independent of weather conditions and environments. However, radar signals can be easily contaminated by various noises such as speckle noise, receiver saturation, and multipath reflection, which can worsen sensing performance. To handle this problem, we propose a learning-free noise removal technique for radar to enhance detection performance. The proposed method leverages adaptive thresholding to remove speckle noise and receiver saturation, and wavelet transform to detect multipath reflection. After noise removal, the radar image is reconstructed with the geometric structure of the surrounding environments. We verify that our method effectively eliminated noise and can be applied to autonomous driving by improving the accuracy of odometry and place recognition.

• 한국측량학회지
• /
• 제32권2호
• /
• pp.133-142
• /
• 2014

Landsat은 대표적인 지구관측 위성 중 하나로 지표면 모니터링, 변화탐지, 분류 등 다양한 분야에서 사용되고 있다. 하지만 구름과 구름의 그림자는 지표의 관측과 분석을 제한하는 장애물 중 하나로, Landsat을 사용하기 전 구름을 제거하고 원래의 지표 피복으로 복원하는 과정은 필수적이다. 최근에 발사된 Landsat-8은 기존위성에 비해 2개의 추가적인 costal/aerosol, cirrus 밴드를 제공하며, 이는 구름을 탐지하고 복원하는데 효율적으로 사용될 수 있다. 따라서 본 연구에서는 Landsat-8의 영상에서 구름을 효과적으로 탐지하고, 복원하는 기법을 단계적으로 제안하였다. Otsu 임계화 기법을 통하여 구름과 구름의 그림자 지역을 탐지하였고, 탐지된 구름 및 그림자 지역은 실험 영상과 참조영상을 이용하여 원래의 지표 피복으로 복원 하였다. 복원영상의 정확도 평가에서는 전체정확도가 약 85%, 카파계수가 0.7128로 본 연구에서 제안한 알고리즘이 효율적임을 확인하였다.

• International Journal of Computer Science & Network Security
• /
• 제24권2호
• /
• pp.158-168
• /
• 2024

There is hardly any person in modern times who has not taken soft drinks instead of drinking water. The rate of people taking soft drinks being surprisingly high, researchers around the world have cautioned from time to time that these drinks lead to weight gain, raise the risk of non-communicable diseases and so on. Therefore, in this work an image-based tool is developed to monitor the nutritional information of soft drinks by using deep convolutional neural network with transfer learning. At first, visual saliency, mean shift segmentation, thresholding and noise reduction technique, collectively known as 'pre-processing' are adopted to extract the location of drinks region. After removing backgrounds and segment out only the desired area from image, we impose Discrete Wavelength Transform (DWT) based resolution enhancement technique is applied to improve the quality of image. After that, transfer learning model is employed for the classification of drinks. Finally, nutrition value of each drink is estimated using Bag-of-Feature (BoF) based classification and Euclidean distance-based ratio calculation technique. To achieve this, a dataset is built with ten most consumed soft drinks in Bangladesh. These images were collected from imageNet dataset as well as internet and proposed method confirms that it has the ability to detect and recognize different types of drinks with an accuracy of 98.51%.

• 대한원격탐사학회지
• /
• 제39권6_1호
• /
• pp.1211-1224
• /
• 2023

구름은 광학위성을 이용한 국토 관측 및 재난 대응, 변화 탐지 등 지표의 현상을 관측하는데 있어 많은 어려운 문제를 야기한다. 구름의 존재는 영상 처리 단계 뿐만 아니라 최종적으로는 데이터의 품질에 영향을 미치므로 이를 반드시 식별하고 제거하는 과정이 필요하다. 따라서 본 연구에서는 위성영상 내 구름의 분광패턴에 가장 근접한 화소를 탐색 및 추출해 최적의 임계값을 선정하고 임계값을 바탕으로 구름 산출물을 제작하는 일련의 과정을 자동으로 수행하는 새로운 구름 탐지 기법을 개발하고자 하였다. 구름 탐지 기법은 크게 세 단계로 구성된다. 첫 번째 단계에서는 Digital Number (DN) 단위 영상을 대기상층 반사율 단위로 변환하는 과정을 수행한다. 두 번째 단계에서는 대기상층 반사율 영상을 이용하여 Hue-Value-Saturation (HSV) 변환 및 삼각형 임계 처리, 최대우도 분류 등의 전처리를 적용하고 각 영상별로 초기 구름 마스크 생성을 위한 임계값을 결정한다. 세번째 후처리 단계에서는 생성된 초기 구름 마스크에 포함된 노이즈를 제거하고 구름 경계 및 내부를 개선한다. 구름 탐지를 위한 실험 자료로 구름의 공간적, 계절적 분포의 다양성을 보여주는 4~11월 시기에 한반도 지역에서 촬영된 국토위성 L2G 영상을 사용하였다. 제안 방법의 성능을 검증하기 위해 단일 임계화 방법으로 생성된 결과를 비교하였다. 실험 결과, 제안 방법은 기존 방법과 비교하여 전처리 과정을 통해 각 영상의 방사학적 특성을 고려할 수 있어 보다 정확하게 구름을 검출할 수 있었다. 또한, 구름 개체를 제외한 나머지 밝은 물체(판넬식 지붕, 콘크리트 도로, 모래 등)의 영향을 최소화하는 결과를 보여주었다. 제안 방법은 기존 방법 대비 F1-score 기준으로 30% 이상의 개선된 결과를 보여주었으나 눈이 포함된 특정 영상에서 한계점이 있었다.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2003년도 ICCAS
• /
• pp.789-792
• /
• 2003

Pipe inspection has a great importance to ensure safety for the nuclear power plant. In this paper, we designed visual inspection module for the tube internal, which diameter is 15${\sim}$20mm. And we made inspection module which consisted of CCD camera and light. And the relation between image and real world coordinate is established. Image processing is performed to calculate mapping parameter and analyze the size of defect. For the calculation of mapping parameter, experiment is performed using grid type test pattern. Acquired image is processed to extract image coordinate. Edge detection, thresholding, median filtering and morphology filtering is applied to extract grid pattern. Extracted image coordinate is used to calculate image to real world mapping. Lens distortion was considered and corrected to get exact data. Coordinate transformation data is provided for the users to recognize easily. Experiment was performed using grid type test pattern, we extracted lens distortion parameter and real coordinate of defect point. Radial distortion of lens was corrected but tangential distortion was not considered. As continuum to this study, the tangential distortion of lens is considered and improvement of analy zing technique for the tube internal be explored continuously.

• Korean Journal of Geomatics
• /
• 제3권1호
• /
• pp.67-72
• /
• 2003

We have seen rapid changes in road systems and networks in urban areas due to fast urbanization and increased traffic demands. As a result, many researchers have put greater importance on extraction, correction and updating of information about road systems. Also, by using the various data on road systems and its condition, we can manage our road more efficiently and economically. Furthermore, such information can be used as input for digital map and GIS analysis. In this research, we used a high resolution aerial photo of the roads in Seongnam area. First, we applied the top-hat filter to the area of interest so that the road markings could be extracted in an efficient manner. The lane separation lines were selected, considering the shape similarity between the selected lane separation line and reference data. Next, we extracted the roads in the urban area using the aforementioned road marking. Using this technique, we could easily extract roads in urban area in semi-automatic way.

• Fisheries and Aquatic Sciences
• /
• 제4권1호
• /
• pp.10-17
• /
• 2001

One serious problem among the troubles to identify objects in an optical microscopic image is contour background due to non-uniform light source and various transparency of samples. To solve this problem, this paper proposed an elimination method of the contour background and compensation technique as follows. First, Otsu's optimal thresholding method extracts pixels representing background. Second, bilinear interpolation finds non-deterministic background pixels among the sampled pixels. Third, the 2D cubic fitting method composes surface function from pivoted background pixels. Fourth, reconstruction procedure makes a contour image from the surface function. Finally, elimination procedure subtracts the approximated background from the original image. To prove the effectiveness of the proposed algorithm, this algorithm is applied to the yeast Zygosaccharomyces rouxii and ammonia-oxidizing bacteria Acinetobacter sp. Labeling by this proposed method can remove some noise and is more exact than labeling by only Otsu's method. Futhermore, we show that it is more effective for the reduction of noise.

• 대한전자공학회논문지SP
• /
• 제38권4호
• /
• pp.422-433
• /
• 2001

본 논문에서는 영상의 그래디언트 (gradient)를 구하여 그래디언트 값의 분포를 마치 3차원 지형과 같은 개념으로 간주하고 여기에 물이 흐르는 개념을 적용한 에지 (edge) 검출 방법을 제안하였다 영상에서 그래디언트 값이 큰 부분은 배경과 객체간의 에지라 볼 수 있으며, 이 에지에 물이 고이게 하기 위해서는 반전된 그래디언트 영상을 사용한다. 반전된 그래디언트 영상에서 물의 흐름을 기반으로 한 enhancing 작업과 국부적응 임계값 적용을 실시하여 잡음을 줄인 에지 영상을 찾는 방법을 제안한다. 합성영상과 실제영상에 대한실험을 통해 제안한 방법의 효율성을 검증하였다.

• 한국정밀공학회지
• /
• 제16권5호통권98호
• /
• pp.48-57
• /
• 1999

Applications of thresholding technique are based on the assumption that object and background pixels in a digital image can be distinguished by their gray level values. For the segmentation of more complex images, it is necessary to resort to multiple threshold selection techniques. This paper describes a new method for multiple threshold selection of gray level images which are not clearly distinguishable from the background. The proposed method consists of three main stages. In the first stage, a probability distribution function for a gray level histogram of an image is derived. Cluster points are defined according to the probability distribution function. In the second stage, fuzzy partition matrix of the probability distribution function is generated through the fuzzy clustering process. Finally, elements of the fuzzy partition matrix are classified as clusters according to gray level values by using max-membership method. Boundary values of classified clusters are selected as multiple threshold. In order to verify the performance of the developed algorithm, automatic inspection process of ball grid array is presented.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제11권11호
• /
• pp.5419-5435
• /
• 2017

The construction of panoramic images on smartphones and low-powered devices is a challenging task. In this paper, we propose a new approach for smoothly stitching images on mobile phones in the presence of moving objects in the scene. Our main contributions include handling moving object problems, reducing processing time, and generating rectangular panoramic images. First, unique and robust feature points are extracted using fast ORB method and a feature matching technique is applied to match the extracted feature points. After obtaining good matched feature points, we employ the non-deterministic RANSAC algorithm to discard wrong matches, and the hommography transformation matrix parameters are estimated with the algorithm. Afterward, we determine precise overlap regions of neighboring images and calculate their absolute differences. Then, thresholding operation and noise removal filtering are applied to create a mask of possible moving object regions. Sequentially, an optimal seam is estimated using dynamic programming algorithm, and a combination of linear blending with the mask information is applied to avoid seam transition and ghosting artifacts. Finally, image-cropping operation is utilized to obtain a rectangular boundary image from the stitched image. Experiments demonstrate that our method is able to produce panoramic images quickly despite the existence of moving objects.