• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.39 no.7
• /
• pp.665-671
• /
• 2015

The performance of fault diagnostics for a planetary gearbox depends on vibration modulation characteristics, which can vary with manufacturing & assembly tolerance, and load condition. In this paper, a fault diagnostics technique that considers vibration modulation characteristics is proposed for the effective fault detection of planetary gearboxes in wind turbines. For identifying the vibration modulation characteristics in practice, re-sampled vibration signals are processed with narrow band-pass filters. Thereafter, the optimal position of the vibration extraction window is identified for effective detection of faulty signals under the varying vibration modulation characteristics. The proposed diagnostics technique makes it possible to perform robust diagnostics of the planetary gearbox with regard to the changeable vibration modulation effect. For demonstrating the proposed fault diagnostics technique, a 2-kW WT testbed is designed with two DC motors and gearboxes. A faulty gear with partial tooth breakage is machined and assembled into the gearbox.

• Korean Journal of Remote Sensing
• /
• v.38 no.6_1
• /
• pp.1231-1244
• /
• 2022

The frequency of exposure of field crops to stress situations is increasing due to abnormal weather conditions. In South Korea, large-scale diseases in representative paddy rice cultivation area were happened. There are limits to field investigation on the crop damage due to large-scale. Satellite-based remote sensing techniques are useful for monitoring crops in cities and counties, but the sensitivity of vegetation index measured from satellite under abnormal growth of crop should be evaluated. The goal is to evaluate satellite-based normalized difference vegetation index (NDVI) retrieved from different spatial scales using drone imagery. In this study, Sentinel-2 and Landsat-8 satellites were used and they have spatial resolution of 10 and 30 m. Drone-based NDVI, which was resampled to the scale of satellite data, had correlation of 0.867-0.940 with Sentinel-2 NDVI and of 0.813-0.934 with Landsat-8 NDVI. When the effects of bias were minimized, Sentinel-2 NDVI had a normalized root mean square error of 0.2 to 2.8% less than that of the drone NDVI compared to Landsat-8 NDVI. In addition, Sentinel-2 NDVI had the constant error values regardless of diseases damage. On the other hand, Landsat-8 NDVI had different error values depending on degree of diseases. Considering the large error at the boundary of agricultural field, high spatial resolution data is more effective in monitoring crops.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2019.10a
• /
• pp.1019-1022
• /
• 2019

최근의 영상 처리 분야는 딥러닝 기법들의 성능이 입증됨에 따라 다양한 분야에서 이와 같은 기법들을 활용해 영상에 대한 분류, 분석, 검출 등을 수행하려는 시도가 활발하다. 그중에서도 의료 진단 보조 역할을 할 수 있는 의료 영상 분석 소프트웨어에 대한 기대가 증가하고 있는데, 본 연구에서는 캡슐내시경 영상에 주목하였다. 캡슐내시경은 주로 소장 촬영을 목표로 하며 식도부터 대장까지 약 8~10시간 동안 촬영된다. 이로 인해 CT, MR, X-ray와 같은 다른 의료 영상과 다르게 하나의 데이터 셋이 10~15만 장의 이미지를 갖는다. 일반적으로 캡슐내시경 영상을 판독하는 순서는 위장관 교차점(Z-Line, 유문판, 회맹판)을 기준으로 위장관 랜드마크(식도, 위, 소장, 대장)를 구분한 뒤, 각 랜드마크 별로 병변 정보를 찾아내는 방식이다. 그러나 워낙 방대한 영상 데이터를 가지기 때문에 의사 혹은 의료 전문가가 영상을 판독하는데 많은 시간과 노력이 소모되고 있다. 본 논문의 목적은 캡슐내시경 영상의 판독에서 모든 환자에 대해 공통으로 수행되고, 판독하는 데 많은 시간을 차지하는 위장관 랜드마크를 찾는 것에 있다. 이를 위해, 위장관 랜드마크를 식별할 수 있는 CNN 학습 모델을 설계하였으며, 더욱 효과적인 학습을 위해 전처리 과정으로 학습에 방해가 되는 학습 노이즈 영상들을 제거하고 위장관 랜드마크 별 특징 분석을 진행하였다. 총 8명의 환자 데이터를 가지고 학습된 모델에 대해 평가 및 검증을 진행하였는데, 무작위로 환자 데이터를 샘플링하여 학습한 모델을 평가한 결과, 평균 정확도가 95% 가 확인되었으며 개별 환자별로 교차 검증 방식을 진행한 결과 평균 정확도 67% 가 확인되었다.

• Smart Media Journal
• /
• v.12 no.1
• /
• pp.9-16
• /
• 2023

Recently, a convolutional neural network (CNN) based system is being developed to overcome the limitations of human resources in the apple quality classification of farmhouse. However, since convolutional neural networks receive only images of the same size, preprocessing such as sampling may be required, and in the case of oversampling, information loss of the original image such as image quality degradation and blurring occurs. In this paper, in order to minimize the above problem, to generate a image patch based graph of an original image and propose a random walk-based positional encoding method to apply the graph transformer model. The above method continuously learns the position embedding information of patches which don't have a positional information based on the random walk algorithm, and finds the optimal graph structure by aggregating useful node information through the self-attention technique of graph transformer model. Therefore, it is robust and shows good performance even in a new graph structure of random node order and an arbitrary graph structure according to the location of an object in an image. As a result, when experimented with 5 apple quality datasets, the learning accuracy was higher than other GNN models by a minimum of 1.3% to a maximum of 4.7%, and the number of parameters was 3.59M, which was about 15% less than the 23.52M of the ResNet18 model. Therefore, it shows fast reasoning speed according to the reduction of the amount of computation and proves the effect.

• Journal of IKEEE
• /
• v.23 no.4
• /
• pp.1250-1256
• /
• 2019

Although one can easily generate real-world 3D mesh data using a 3D printer or a depth camera, the generated data inevitably includes unnecessary noise. Therefore, mesh denoising is essential to obtain intact 3D mesh data. However, conventional mathematical denoising methods require preprocessing and often eliminate some important features of the 3D mesh. To address this problem, this paper proposes a deep learning based 3D mesh denoising method. Specifically, we propose a convolution-based autoencoder model consisting of an encoder and a decoder. The convolution operation applied to the mesh data performs denoising considering the relationship between each vertex constituting the mesh data and the surrounding vertices. When the convolution is completed, a sampling operation is performed to improve the learning speed. Experimental results show that the proposed autoencoder model produces faster and higher quality denoised data than the conventional methods.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.51 no.8
• /
• pp.87-96
• /
• 2014

In this paper, we propose a directionally adaptive aliasing and noise removal using dictionary learning based on space-frequency analysis. The proposed aliasing and noise removal algorithm consists of two modules; i) aliasing and noise detection using dictionary learning and analysis of frequency characteristics from the combined wavelet-Fourier transform and ii) aliasing removal with suppressing noise based on the directional shrinkage in the detected regions. The proposed method can preserve the high-frequency details because aliasing and noise region is detected. Experimental results show that the proposed algorithm can efficiently reduce aliasing and noise while minimizing losses of high-frequency details and generation of artifacts comparing with the conventional methods. The proposed algorithm is suitable for various applications such as image resampling, super-resolution image, and robot vision.

• The Journal of Bigdata
• /
• v.8 no.1
• /
• pp.157-165
• /
• 2023

Seafood is a major source of protein in many countries and its consumption is increasing. In Korea, consumption of seafood is increasing, but self-sufficiency rate is decreasing, and the importance of safety management is increasing as the amount of imported seafood increases. There are hundreds of species of aquatic products imported into Korea from over 110 countries, and there is a limit to relying only on the experience of inspectors for safety management of imported aquatic products. Based on the data, a model that can predict the customs inspection results of imported aquatic products is developed, and a machine learning classification model that determines the non-conformity of aquatic products when an import declaration is submitted is created. As a result of customs inspection of imported marine products, the nonconformity rate is less than 1%, which is very low imbalanced data. Therefore, a sampling method that can complement these characteristics was comparatively studied, and a preprocessing method that can interpret the classification result was applied. Among various machine learning-based classification models, Random Forest and XGBoost showed good performance. The model that predicts both compliance and non-conformance well as a result of the clearance inspection is the basic random forest model to which ADASYN and one-hot encoding are applied, and has an accuracy of 99.88%, precision of 99.87%, recall of 99.89%, and AUC of 99.88%. XGBoost is the most stable model with all indicators exceeding 90% regardless of oversampling and encoding type.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.40 no.5
• /
• pp.312-321
• /
• 2003

In this paper, we implement robot which are ability to recognize obstacles and moving automatically to destination. we present two results in this paper; hardware implementation of image processing board and software implementation of visual feedback algorithm for a self-controlled robot. In the first part, the mobile robot depends on commands from a control board which is doing image processing part. We have studied the self controlled mobile robot system equipped with a CCD camera for a long time. This robot system consists of a image processing board implemented with DSPs, a stepping motor, a CCD camera. We will propose an algorithm in which commands are delivered for the robot to move in the planned path. The distance that the robot is supposed to move is calculated on the basis of the absolute coordinate and the coordinate of the target spot. And the image signal acquired by the CCD camera mounted on the robot is captured at every sampling time in order for the robot to automatically avoid the obstacle and finally to reach the destination. The image processing board consists of DSP (TMS320VC33), ADV611, SAA7111, ADV7l76A, CPLD(EPM7256ATC144), and SRAM memories. In the second part, the visual feedback control has two types of vision algorithms: obstacle avoidance and path planning. The first algorithm is cell, part of the image divided by blob analysis. We will do image preprocessing to improve the input image. This image preprocessing consists of filtering, edge detection, NOR converting, and threshold-ing. This major image processing includes labeling, segmentation, and pixel density calculation. In the second algorithm, after an image frame went through preprocessing (edge detection, converting, thresholding), the histogram is measured vertically (the y-axis direction). Then, the binary histogram of the image shows waveforms with only black and white variations. Here we use the fact that since obstacles appear as sectional diagrams as if they were walls, there is no variation in the histogram. The intensities of the line histogram are measured as vertically at intervals of 20 pixels. So, we can find uniform and nonuniform regions of the waveforms and define the period of uniform waveforms as an obstacle region. We can see that the algorithm is very useful for the robot to move avoiding obstacles.