• International Journal of Computer Science & Network Security
• /
• 제22권6호
• /
• pp.364-373
• /
• 2022

Image morphing methods make seamless transition changes in the image and mask the meaningful information attached to it. This can be detected by traditional machine learning algorithms and new emerging deep learning algorithms. In this research work, scope of different Hybrid learning approaches having combination of Deep learning and Machine learning are being analyzed with the public dataset CASIA V1.0, CASIA V2.0 and DVMM to find the most efficient algorithm. The simulated results with CNN (Convolution Neural Network), Hybrid approach of CNN along with SVM (Support Vector Machine) and Hybrid approach of CNN along with Random Forest algorithm produced 96.92 %, 95.98 and 99.18 % accuracy respectively with the CASIA V2.0 dataset having 9555 images. The accuracy pattern of applied algorithms changes with CASIA V1.0 data and DVMM data having 1721 and 1845 set of images presenting minimal accuracy with Hybrid approach of CNN and Random Forest algorithm. It is confirmed that the choice of best algorithm to find image forgery depends on input data type. This paper presents the combination of best suited algorithm to detect image morphing with different input datasets.

• 대한원격탐사학회지
• /
• 제39권3호
• /
• pp.269-282
• /
• 2023

본 연구는 딥러닝 모델(deep learning model)을 활용하여 토지피복분류를 수행하였으며 입력 이미지의 크기, Stride 적용 등 데이터세트(dataset)의 조절을 통해 토지피복분류를 위한 최적의 딥러닝 모델 선정을 목적으로 하였다. 적용한 딥러닝 모델은 3종류로 Encoder-Decoder 구조를 가진 U-net과 DeeplabV3+, 두 가지 모델을 결합한 앙상블(Ensemble) 모델을 활용하였다. 데이터세트는 RapidEye 위성영상을 입력영상으로, 라벨(label) 이미지는 Intergovernmental Panel on Climate Change 토지이용의 6가지 범주에 따라 구축한 Raster 이미지를 참값으로 활용하였다. 딥러닝 모델의 정확도 향상을 위해 데이터세트의 질적 향상 문제에 대해 주목하였으며 딥러닝 모델(U-net, DeeplabV3+, Ensemble), 입력 이미지 크기(64 × 64 pixel, 256 × 256 pixel), Stride 적용(50%, 100%) 조합을 통해 12가지 토지피복도를 구축하였다. 라벨 이미지와 딥러닝 모델 기반의 토지피복도의 정합성 평가결과, U-net과 DeeplabV3+ 모델의 전체 정확도는 각각 최대 약 87.9%와 89.8%, kappa 계수는 모두 약 72% 이상으로 높은 정확도를 보였으며, 64 × 64 pixel 크기의 데이터세트를 활용한 U-net 모델의 정확도가 가장 높았다. 또한 딥러닝 모델에 앙상블 및 Stride를 적용한 결과, 최대 약 3% 정확도가 상승하였으며 Semantic Segmentation 기반 딥러닝 모델의 단점인 경계간의 불일치가 개선됨을 확인하였다.

• 전기전자학회논문지
• /
• 제23권1호
• /
• pp.112-119
• /
• 2019

본 논문에서는 불꽃 감지를 위한 임베디드 시스템에 적합한 딥러닝 구조를 제안한다. 제안하는 딥러닝 구조의 불꽃 감지 과정은 불꽃 색깔 모델을 사용한 불꽃 영역 검출, 불꽃 색깔 특화 딥러닝 구조를 사용한 불꽃 영상 분류, 검출된 불꽃 영역의 $N{\times}N$ 셀 분리, 불꽃 모양 특화 딥러닝 구조를 사용한 불꽃 영상 분류 등의 4가지 과정으로 구성된다. 첫 번째로 입력 영상에서 불꽃의 색만을 추출한 다음 레이블링하여 불꽃 영역을 검출한다. 두 번째로 검출된 불꽃 영역을 불꽃 색깔에 특화 학습된 딥러닝 구조의 입력으로 넣고, 출력단의 불꽃 클래스 확률이 75% 이상에서만 불꽃 영상으로 분류한다. 세 번째로 앞 단에서 75% 미만 불꽃 영상으로 분류된 영상들의 검출된 불꽃 영역을 $N{\times}N$ 단위로 분할한다. 네 번째로 $N{\times}N$ 단위로 분할된 작은 셀들을 불꽃의 모양에 특화 학습된 딥러닝 구조의 입력으로 넣고, 각 셀의 불꽃 여부를 판단하여 50% 이상의 셀들이 불꽃 영상으로 분류될 경우에 불꽃 영상으로 분류한다. 제안된 딥러닝 구조의 성능을 평가하기 위하여 ImageNet의 불꽃 데이터베이스를 사용하여 실험하였다. 실험 결과, 제안하는 딥러닝 구조는 기존의 딥러닝 구조보다 평균 29.86% 낮은 리소스 점유율과 8초 빠른 불꽃 감지 시간을 나타내었다. 불꽃 검출률은 기존의 딥러닝 구조와 비교하여 평균 0.95% 낮은 결과를 나타내었으나, 이는 임베디드 시스템에 적용하기 위해 딥러닝 구조를 가볍게 구성한데서 나온 결과이다. 따라서 본 논문에서 제안하는 불꽃 감지를 위한 딥러닝 구조는 임베디드 시스템 적용에 적합함이 입증되었다.

• 한국음향학회지
• /
• 제42권4호
• /
• pp.370-376
• /
• 2023

수중영상은 수중 잡음과 낮은 해상도로 표적의 형상과 구분이 명확하지 않다. 그리고 딥러닝의 입력으로 수중영상은 전처리가 필요하며 Segmentation이 선행되어야 한다. 전처리를 하여도 표적은 명확하지 않으며 딥러닝에 의한 탐지, 식별의 성능도 높지 않을 수 있다. 따라서 표적을 구분하며 명확하게 하는 작업이 필요하다. 본 연구에서는 수중영상에서 표적 그림자의 중요성을 확인하고 그림자에 의한 물체 탐지 및 표적 영역 획득, 그리고 수중배경이 없는 표적과 그림자만의 형상이 담긴 데이터를 생성하며 더 나아가 픽셀값이 일정하지 않은 표적과 그림자 영상을 표적은 흰색, 그림자는 흑색, 그리고 배경은 회색의 3-모드의 영상으로 변환하는 과정을 제시한다. 이를 통해 딥러닝의 입력으로 명확히 전처리된 판별이 용이한 영상을 제공할 수 있다. 또한 처리는 Open Source Computer Vision(OpenCV)라이브러리의 영상처리 코드를 사용했으면 처리 속도도 역시 실시간 처리에 적합한 결과를 얻었다.

• 천문학회보
• /
• 제44권2호
• /
• pp.42.3-42.3
• /
• 2019

We translate Solar Dynamics Observatory/Atmospheric Imaging Assembly (AIA) ultraviolet (UV) multi-channel images into another UV single-channel image using a deep learning algorithm based on conditional generative adversarial networks (cGANs). The base input channel, which has the highest correlation coefficient (CC) between UV channels of AIA, is 193 Å. To complement this channel, we choose two channels, 1600 and 304 Å, which represent upper photosphere and chromosphere, respectively. Input channels for three models are single (193 Å), dual (193+1600 Å), and triple (193+1600+304 Å), respectively. Quantitative comparisons are made for test data sets. Main results from this study are as follows. First, the single model successfully produce other coronal channel images but less successful for chromospheric channel (304 Å) and much less successful for two photospheric channels (1600 and 1700 Å). Second, the dual model shows a noticeable improvement of the CC between the model outputs and Ground truths for 1700 Å. Third, the triple model can generate all other channel images with relatively high CCs larger than 0.89. Our results show a possibility that if three channels from photosphere, chromosphere, and corona are selected, other multi-channel images could be generated by deep learning. We expect that this investigation will be a complementary tool to choose a few UV channels for future solar small and/or deep space missions.

• 한국정보기술학회 영문논문지
• /
• 제10권2호
• /
• pp.61-72
• /
• 2020

Recently, researches that are using deep learning technology in various fields are being conducted. The fields include geographic map processing. In this paper, I propose a method to infer where the map area included in the image is. The proposed method generates and learns images including a map, detects map areas from input images, extracts character strings belonging to those map areas, and converts the extracted character strings into coordinates through geocoding to infer the coordinates of the input image. Faster R-CNN was used for learning and map detection. In the experiment, the difference between the center coordinate of the map on the test image and the center coordinate of the detected map is calculated. The median value of the results of the experiment is 0.00158 for longitude and 0.00090 for latitude. In terms of distance, the difference is 141m in the east-west direction and 100m in the north-south direction.

• 한국컴퓨터정보학회논문지
• /
• 제23권11호
• /
• pp.43-49
• /
• 2018

In this paper, we propose a displacement measurement method based on deep learning using image data obtained from tensile tests of a material specimen. We focus on the fact that the sequential images during the tension are generated and the displacement of the specimen is represented in the image data. So, we designed sample generation model which makes sequential images of specimen. The behavior of generated images are similar to the real specimen images under tensile force. Using generated images, we trained and validated our model. In the deep neural network, sequential images are assigned to a multi-channel input to train the network. The multi-channel images are composed of sequential images obtained along the time domain. As a result, the neural network learns the temporal information as the images express the correlation with each other along the time domain. In order to verify the proposed method, we conducted experiments by comparing the deformation measuring performance of the neural network changing the displacement range of images.

• International journal of advanced smart convergence
• /
• 제12권2호
• /
• pp.56-66
• /
• 2023

The goal of deep learning is to extract complex features from multidimensional data use the features to create models that connect input and output. Deep learning is a process of learning nonlinear features and functions from complex data, and the user data that is employed to train deep learning models has become the focus of privacy concerns. Companies that collect user's sensitive personal information, such as users' images and voices, own this data for indefinite period of times. Users cannot delete their personal information, and they cannot limit the purposes for which the data is used. The study has designed a deep learning method that employs privacy protection technology that uses distributed collaborative learning so that multiple participants can use neural network models collaboratively without sharing the input datasets. To prevent direct leaks of personal information, participants are not shown the training datasets during the model training process, unlike traditional deep learning so that the personal information in the data can be protected. The study used a method that can selectively share subsets via an optimization algorithm that is based on modified distributed stochastic gradient descent, and the result showed that it was possible to learn with improved learning accuracy while protecting personal information.

• 한국방송∙미디어공학회:학술대회논문집
• /
• 한국방송∙미디어공학회 2019년도 하계학술대회
• /
• pp.98-101
• /
• 2019

At present, deep convolutional neural networks have made a very important contribution in single-image super-resolution. Through the learning of the neural networks, the features of input images are transformed and combined to establish a nonlinear mapping of low-resolution images to high-resolution images. Some previous methods are difficult to train and take up a lot of memory. In this paper, we proposed a simple and compact deep recursive residual network learning the features for single image super resolution. Global residual learning and local residual learning are used to reduce the problems of training deep neural networks. And the recursive structure controls the number of parameters to save memory. Experimental results show that the proposed method improved image qualities that occur in previous methods.

• 대한의용생체공학회:의공학회지
• /
• 제42권4호
• /
• pp.167-174
• /
• 2021

The purpose of this study is to classify TIFF images, PNG images, and JPEG images using deep learning, and to compare the accuracy by verifying the classification performance. The TIFF, PNG, and JPEG images converted from chest X-ray DICOM images were applied to five deep neural network models performed in image recognition and classification to compare classification performance. The data consisted of a total of 4,000 X-ray images, which were converted from DICOM images into 16-bit TIFF images and 8-bit PNG and JPEG images. The learning models are CNN models - VGG16, ResNet50, InceptionV3, DenseNet121, and EfficientNetB0. The accuracy of the five convolutional neural network models of TIFF images is 99.86%, 99.86%, 99.99%, 100%, and 99.89%. The accuracy of PNG images is 99.88%, 100%, 99.97%, 99.87%, and 100%. The accuracy of JPEG images is 100%, 100%, 99.96%, 99.89%, and 100%. Validation of classification performance using test data showed 100% in accuracy, precision, recall and F1 score. Our classification results show that when DICOM images are converted to TIFF, PNG, and JPEG images and learned through preprocessing, the learning works well in all formats. In medical imaging research using deep learning, the classification performance is not affected by converting DICOM images into any format.