• Progress in Medical Physics
• /
• v.10 no.1
• /
• pp.47-54
• /
• 1999

Purpose: To verify the BRW coordinates of target located within the limit of XKnife hardware, and to verify the successful transfer of image data, rod detection, anatomical structure when CT images are transferred into a XKnife computer. Materials and Methods: Target coordinates of 13 patients were calculated by SCS1 computer through the rod image on the console screen and film. BRW coordinates of target and landmark calculated by SCS1 computer were compared to those acquired by XKnife localizer. Results : Vertical components of BRW coordinates of target for 13 patients are larger than -50 mm, and then the vertical components of BRW coordinates of target are localized within the limit of XKnife hardware. Average differences between XKnife and SCS1 for BRW coordinates of target and landmark were within 1 mm for AP and LAT components, 0.5 mm for VERT component. Conclusion : It was verified that the SCS1 computer is adequate tool to calculate BRW coordinates of target quickly. And by the comparison between SCS1 computer and XKnife localizer, it was verified that the image transfer into the XKnife computer was performed successfully.

• Journal of Internet Computing and Services
• /
• v.10 no.5
• /
• pp.183-194
• /
• 2009

In this paper, we propose a new method for detecting objectionable images with an active shape model. Our method first learns the shape of breast lines through principle component analysis and alignment as well as the distribution of intensity values of corresponding landmarks, and then extracts breast lines with the learned shape and intensity distribution. To accurately select the initial position of active shape model, we obtain parameters on scale, rotation, and translation. After positioning the initial location of active shape model using scale and rotation information, iterative searches are performed. We can identify adult images by calculating the average of the distance between each landmark and a candidate breast line. The experiment results show that the proposed method can detect adult images effectively by comparing various results.

• The korean journal of orthodontics
• /
• v.51 no.2
• /
• pp.77-85
• /
• 2021

Objective: To evaluate the accuracy of a multi-stage convolutional neural network (CNN) model-based automated identification system for posteroanterior (PA) cephalometric landmarks. Methods: The multi-stage CNN model was implemented with a personal computer. A total of 430 PA-cephalograms synthesized from cone-beam computed tomography scans (CBCT-PA) were selected as samples. Twenty-three landmarks used for Tweemac analysis were manually identified on all CBCT-PA images by a single examiner. Intra-examiner reproducibility was confirmed by repeating the identification on 85 randomly selected images, which were subsequently set as test data, with a two-week interval before training. For initial learning stage of the multi-stage CNN model, the data from 345 of 430 CBCT-PA images were used, after which the multi-stage CNN model was tested with previous 85 images. The first manual identification on these 85 images was set as a truth ground. The mean radial error (MRE) and successful detection rate (SDR) were calculated to evaluate the errors in manual identification and artificial intelligence (AI) prediction. Results: The AI showed an average MRE of 2.23 ± 2.02 mm with an SDR of 60.88% for errors of 2 mm or lower. However, in a comparison of the repetitive task, the AI predicted landmarks at the same position, while the MRE for the repeated manual identification was 1.31 ± 0.94 mm. Conclusions: Automated identification for CBCT-synthesized PA cephalometric landmarks did not sufficiently achieve the clinically favorable error range of less than 2 mm. However, AI landmark identification on PA cephalograms showed better consistency than manual identification.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• v.13 no.2
• /
• pp.133-139
• /
• 2013

This paper proposes a vision-based indoor localization method for autonomous vehicles. A single upward-facing digital camera was mounted on an autonomous vehicle and used as a vision sensor to identify artificial landmarks and any natural corner features. An interest point detector was used to find the natural features. Using an optical flow detection algorithm, information related to the direction and vehicle translation was defined. This information was used to track the vehicle movements. Random noise related to uneven light disrupted the calculation of the vehicle translation. Thus, to estimate the vehicle translation, a Kalman filter was used to calculate the vehicle position. These algorithms were tested on a vehicle in a real environment. The image processing method could recognize the landmarks precisely, while the Kalman filter algorithm could estimate the vehicle's position accurately. The experimental results confirmed that the proposed approaches can be implemented in practical situations.

• Journal of Korea Multimedia Society
• /
• v.20 no.11
• /
• pp.1720-1728
• /
• 2017

Active shape model is widely used in the field of image processing especially on arbitrary meaningful shape extraction from single gray level image. Cootes et. al. showed efficient detection of variable shape from image by using covariance and mean shape from learning. There are two stages of learning and testing. Hahn applied enhanced shape alignment method rather than using Cootes's rotation and scale scheme. Hahn did not modified the profile itself. In this paper, the method using directional one dimensional profile is proposed to enhance Cootes's one dimensional profile and the shape alignment algorithm of Hahn is combined. The performance of the proposed method was superior to Cootes's and Hahn's. Average landmark estimation error for each image was 27.72 pixels and 39.46 for Cootes's and 33.73 for Hahn's each.

• Proceedings of the IEEK Conference
• /
• 2007.07a
• /
• pp.401-402
• /
• 2007

In this paper, we propose an automatic method to finding corresponding points. One 2D image can be changed 3D shape by 3D model. The main idea is using gabor wavelet values from 3D model. And Elastic Bunch Graph Matching algorithm is more stable in 3D model.

• Proceedings of the Korean Information Science Society Conference
• /
• 2006.06b
• /
• pp.127-129
• /
• 2006

모바일 디바이스에서 얻을 수 있는 로그에는 다양한 개인정보가 풍부하게 포함되어 있으면서도 제약이 많아 활용이 어렵다. 그 동안은 모바일 장치의 용량, 파워의 제약과 정보 분석의 어려움으로 로그 정보를 무시해온 것이 일반적이었다. 본 논문에서는 모바일 디바이스의 다양한 로그 정보를 분석하여 사용자에게 의미 있는 상황(특이성)을 탐지해낼 수 있는 정보 분석 방법을 제안한다. 불확실한 상황에서의 정확성 향상을 위해 규칙/패턴 분석에 의한 특이성 추론뿐만 아니라 베이지안 네트워크를 활용한 확률적인 접근 방법을 활용한다. 이때, 복잡하지 않고 연산이 효율적으로 이루어질 수 있도록 BN을 모듈화하고 모듈화된 BN의 상호보완적인 확률 추론을 위한 BN 처리 과정을 제안한다. 그리고, 특이성 추출 모듈을 주기적으로 업데이트함으로써 성능을 향상시키기 위한 학습알고리즘을 소개한다.

• Korean Journal of Remote Sensing
• /
• v.23 no.4
• /
• pp.297-309
• /
• 2007

The first Korean geostationary weather satellite, Communications, Oceanography and Meteorology Satellite (COMS) will be launched in 2008. The ground station for COMS needs to perform geometric correction to improve accuracy of satellite image data and to broadcast geometrically corrected images to users within 30 minutes after image acquisition. For such a requirement, we developed automated and fast geometric correction techniques. For this, we generated control points automatically by matching images against coastline data and by applying a robust estimation called RANSAC. We used GSHHS (Global Self-consistent Hierarchical High-resolution Shoreline) shoreline database to construct 211 landmark chips. We detected clouds within the images and applied matching to cloud-free sub images. When matching visible channels, we selected sub images located in day-time. We tested the algorithm with GOES-9 images. Control points were generated by matching channel 1 and channel 2 images of GOES against the 211 landmark chips. The RANSAC correctly removed outliers from being selected as control points. The accuracy of sensor models established using the automated control points were in the range of $1{\sim}2$ pixels. Geometric correction was performed and the performance was visually inspected by projecting coastline onto the geometrically corrected images. The total processing time for matching, RANSAC and geometric correction was around 4 minutes.

• The korean journal of orthodontics
• /
• v.54 no.1
• /
• pp.48-58
• /
• 2024

Objective: To quantify the effects of midline-related landmark identification on midline deviation measurements in posteroanterior (PA) cephalograms using a cascaded convolutional neural network (CNN). Methods: A total of 2,903 PA cephalogram images obtained from 9 university hospitals were divided into training, internal validation, and test sets (n = 2,150, 376, and 377). As the gold standard, 2 orthodontic professors marked the bilateral landmarks, including the frontozygomatic suture point and latero-orbitale (LO), and the midline landmarks, including the crista galli, anterior nasal spine (ANS), upper dental midpoint (UDM), lower dental midpoint (LDM), and menton (Me). For the test, Examiner-1 and Examiner-2 (3-year and 1-year orthodontic residents) and the Cascaded-CNN models marked the landmarks. After point-to-point errors of landmark identification, the successful detection rate (SDR) and distance and direction of the midline landmark deviation from the midsagittal line (ANS-mid, UDM-mid, LDM-mid, and Me-mid) were measured, and statistical analysis was performed. Results: The cascaded-CNN algorithm showed a clinically acceptable level of point-to-point error (1.26 mm vs. 1.57 mm in Examiner-1 and 1.75 mm in Examiner-2). The average SDR within the 2 mm range was 83.2%, with high accuracy at the LO (right, 96.9%; left, 97.1%), and UDM (96.9%). The absolute measurement errors were less than 1 mm for ANS-mid, UDM-mid, and LDM-mid compared with the gold standard. Conclusions: The cascaded-CNN model may be considered an effective tool for the auto-identification of midline landmarks and quantification of midline deviation in PA cephalograms of adult patients, regardless of variations in the image acquisition method.

• Journal of Advanced Navigation Technology
• /
• v.24 no.5
• /
• pp.430-437
• /
• 2020

The proposed model implements a model that improves the face prediction rate and recognition rate through learning with an artificial neural network using face detection, landmark and face recognition algorithms. After landmarking in the face images of a specific person, the proposed model use the previously learned Caffe model to extract face detection and embedding vector 128D. The learning is learned by building machine learning algorithms such as support vector machine (SVM) and deep neural network (DNN). Face recognition is tested with a face image different from the learned figure using the learned model. As a result of the experiment, the result of learning with DNN rather than SVM showed better prediction rate and recognition rate. However, when the hidden layer of DNN is increased, the prediction rate increases but the recognition rate decreases. This is judged as overfitting caused by a small number of objects to be recognized. As a result of learning by adding a clear face image to the proposed model, it is confirmed that the result of high prediction rate and recognition rate can be obtained. This research will be able to obtain better recognition and prediction rates through effective deep learning establishment by utilizing more face image data.