본문 바로가기 주메뉴 바로가기
국회도서관 홈으로 정보검색 소장정보 검색

초록보기

MRI(Magnetic Resonance Imaging) 영상과 CT(Computed Tomography) 영상과 같은 의료영상에서 잡음제거는 의료영상 시스템의 성능에 중요한 영향을 미친다. 최근 영상처리 기술에 딥러닝(Deep Learning)의 도입으로 잡음제거 방법들의 성능이 향상되고 있다. 그러나 영상영역에서 디테일을 보존하면서 잡음만을 제거하는 것은 한계가 있다. 본 논문에서는 웨이블렛 변환 기반 CNN(Convolutional Neural Network) 모형, 즉 WT-DnCNN(Wavelet Transform-Denoising Convolutional Neural Network) 모형을 통해 잡음제거 성능을 높이고자 한다. 이는 잡음 영상에 웨이블렛 변환을 사용하여 주파수 대역별로 구분하여 일차적으로 잡음을 제거하고, 해당 주파수 대역에서 기존 DnCNN 모형을 적용하여 최종적으로 잡음을 제거하고자 한다. 본 논문에서 제안된 WT-DnCNN 모형의 성능평가를 위해 다양한 잡음, 즉, 가우시안 잡음(Gaussian Noise), 포아송 잡음(Poisson Noise) 그리고 스펙클 잡음(Speckle Noise)에 의해 훼손된 MRI 영상과 CT 영상을 대상으로 실험하였다. 성능 실험 결과, WT-DnCNN 모형은 정성적 비교에서 전통적인 필터 즉, BM3D(Block-Matching and 3D Filtering) 필터뿐만 아니라 기존의 딥러닝 모형인 DnCNN, CDAE(Convolution Denoising AutoEncoder) 모형보다 우수하고, 정량적 비교에서 PSNR(Peak Signal-to-Noise Ratio) 과 SSIM(Structural Similarity Index Measure) 수치는 MRI 영상에서 각각 36~43과 0.93~0.98, CT 영상에서 각각 38~43과 0.95~0.98 정도로 우수한 결과를 보였다. 또한, 모형의 실행 속도 비교에서 DnCNN 모형은 BM3D 모형보다는 훨씬 적게 결렸으나 DnCNN 모형과의 비교에서는 웨이블렛 변환 추가로 인해 오래 걸림을 알 수 있었다.

In medical images such as MRI(Magnetic Resonance Imaging) and CT(Computed Tomography) images, noise removal has a significant impact on the performance of medical imaging systems. Recently, the introduction of deep learning in image processing technology has improved the performance of noise removal methods. However, there is a limit to removing only noise while preserving details in the image domain. In this paper, we propose a wavelet transform-based CNN(Convolutional Neural Network) model, namely the WT-DnCNN(Wavelet Transform-Denoising Convolutional Neural Network) model, to improve noise removal performance. This model first removes noise by dividing the noisy image into frequency bands using wavelet transform, and then applies the existing DnCNN model to the corresponding frequency bands to finally remove noise. In order to evaluate the performance of the WT-DnCNN model proposed in this paper, experiments were conducted on MRI and CT images damaged by various noises, namely Gaussian noise, Poisson noise, and speckle noise. The performance experiment results show that the WT-DnCNN model is superior to the traditional filter, i.e., the BM3D(Block-Matching and 3D Filtering) filter, as well as the existing deep learning models, DnCNN and CDAE(Convolution Denoising AutoEncoder) model in qualitative comparison, and in quantitative comparison, the PSNR(Peak Signal-to-Noise Ratio) and SSIM(Structural Similarity Index Measure) values were 36~43 and 0.93~0.98 for MRI images and 38~43 and 0.95~0.98 for CT images, respectively. In addition, in the comparison of the execution speed of the models, the DnCNN model was much less than the BM3D model, but it took a long time due to the addition of the wavelet transform in the comparison with the DnCNN model.

권호기사

권호기사 목록 테이블로 기사명, 저자명, 페이지, 원문, 기사목차 순으로 되어있습니다.
기사명 저자명 페이지 원문 목차
Comparative analysis of RNN architectures and activation functions with attention mechanisms for mars weather prediction Jaehyeok Jo, Yunho Sin, Bo-Young Kim, Jihoon Moon p. 1-9

Research on constructing a sentiment lexicon for the F&B sector based on the N-gram framework Yeryung Moon, Gaeun Son, Geonuk Nam, Hanjin Lee p. 11-19

Medical image denoising using wavelet transform-based CNN model Seoyun Jang, Dong Hoon Lim p. 21-34

3D object detection via multi-scale feature knowledge distillation Se-Gwon Cheon, Hyuk-Jin Shin, Seung-Hwan Bae p. 35-45

Multiple camera-based real-time long queue vision algorithm for public safety and efficiency Tae-hoon Kim, Ji-young Na, Ji-won Yoon, Se-Hun Lee, Jun-ho Ahn p. 47-57

Radio frequency-based drone detection and classification using discrete Fourier transform and LightGBM Ki-Hyeon Sung, Soo-Jin Lee p. 59-68

(A) study on the impact of noise on YOLO-based object detection in autonomous driving environments Ra Yeong Kim, Hyun-Jong Cha, Ah Reum Kang p. 69-75

(A) study on measuring the risk of re-identification of personal information in conversational text data using AI Dong-Hyun Kim, Ye-Seul Cho, Tae-Jong Kim p. 77-87

Real-time defog processing using cooperative networks Sanghyun Jung p. 89-96

(The) method for classifying stainless steel grades in products using portable NIR spectrometer and CNN Ju-Hoon Jang, In-Yeop Choi p. 97-104

Named entity normalization for traditional herbal formula mentions Ho Jang p. 105-111

Exploring predictive models for student success in national physical therapy examination : machine learning approach Bokyung Kim, Yeonseop Lee, Jang-hoon Shin, Yusung Jang, Wansuk Choi p. 113-120

(A) temporal convolutional network for hotel demand prediction based on NSGA3 feature selection Keehyun Park, Gyeongho Jung, Hyunchul Ahn p. 121-128

(A) design and implementation of a worker musculoskeletal assessment platform based on machine learning Sejong Lee p. 129-135

(A) YOLOv8-based two-stage framework for non-destructive detection of Varroa destructor infestations in Apis mellifera colonies Yongsun Lee, Hyunsu Cho, Bo-Young Kim, Jihoon Moon p. 137-148

Implementation of a micro drill bit foreign matter inspection system using deep learning Jung-Sub Kim, Tae-Sung Kim, Gyu-Seok Lee p. 149-156

(A) particle-grid method for efficient sound synthesis of ocean waves Jong-Hyun Kim p. 157-164

Towards next generation game development : a comprehensive analysis of game engines technologies Soo Kyun Kim, Iqbal Muhamad Ali, Min Woo Ha p. 165-173

(An) empirical study on visitors’ experiences with the <Game Society> Bo-A Rhee, You-min Im, So-Eun Park p. 175-184

(A) study on visitor engagement of the audio guide and curating-bot at museum of modern and contemporary art Bo-A Rhee, Ye-Eun Seo, Yeon-U Ro, Gyu-Hyun Kim p. 185-195

Research perspectives and research issues on the origins of Korean technology culture : focusing on amateur radio Sangkil Yoon p. 197-205

Development of a model for identifying drug organizations and their scale through tweet clustering Jin-Gyeong Kim, Eun-Young Park, Da-Sol Kim, Cho-Won Kim, Jiyeon Kim p. 207-218

(A) library for object-to-graph mapping with annotations in Java Ji-Woong Choi p. 219-228

(The) impact of disability acceptance on life satisfaction : focusing on the mediating effect of self-efficacy and the moderating effect of the degree of participation in social activities Hyung-Hee Kim p. 229-236

(The) effect of cultural values between countries on the entrepreneurial cognition and attitude Xu Haichao, Yunkyung Lee, Youngsik Kwak p. 237-243

Analysis of the relationship between university facilities and student satisfaction with campus life using structural equation modeling : verification of the mediating effect of university and department pride Jae-Nam Kim, Seong-Eun Chu p. 245-253

Effect of a career education program using portfolio on career maturity and self-efficacy of female college students Jung-chul Heo p. 255-264

(The) relationship between streaming hits view and sales performance in live streaming commerce Jong-Hee Jeong, Jae-Ik Shin p. 265-274

(The) effect of artificial intelligence-based educational app on writing ability of first graders of elementary school Sojeong Choi, Kwihoon Kim p. 275-286