|
ROI-based Fully Automated Liver Registration in Multi-phase CT Images |
|
Kentaro Saito, Hyoungseop Kim(Kyushu institute of technology, Japan) |
| We propose a registration method for fully automated liver tumor detection in Multi-phase CT Images. Registration accuracy is important when obtaining image features from multiple time phases. we propose a robust initial alignment method independent of changing image density features in each time phase, and deformable registration method with region of interests as liver region extracted by U-Net. Experimental results show that segmentation of early arterial phase is 83% and registration is 93% accuracy. |
|
|
Comparative Study of Character Recognition on Thai License Plate using DCT and FIR System |
|
Kaset Sirisantisamrid, Napasool Wongvanich, Suphan Gulpanich(King Mongkut's Institute of Technology Ladkrabang, Thailand) |
| The recognition of characters and numbers on Thai license plate using DCT and FIR system methods is proposed. For DCT method, there are two ways: 1) compute the coefficients of 1D-DCT by separation in the complements of character 2) compute the power spectrum of DCT and use them as the character features. In FIR system method, the impulse responses of FIR system are use as the features of character. The 120 car images under the different of light conditions are used to testing all three. In the results, the successful of recognition rate for separation of DCT coefficients, power spectrum of DC |
|
|
Recognition of Single-Land Countries on Outline Images by Using BAS Feature |
|
Eren Yıldırım(Bahcesehir University, Korea), Ömer Faruk Ince(Kyungsung University, Korea), Yucel Batu Salman, Ege Sadıc(Bahcesehir University, Turkey), Jang Sik Park(Kyungsung University, Korea) |
| This paper presents an approach for the recognition of single-land countries on outline satellite images by using Beam Angle Statistics (BAS). After converting the RGB images to binary and applying the preprocessing steps, single contour is extracted by Canny Edge detector and BAS
feature is extracted from contour point vector. Each country has unique feature vector and is independent from scale and rotation. Tests were conducted between the
proposed algorithm, SIFT and ORB features. Results
show that BAS outperforms other descriptors. |
|
|
Ensemble Grid Formation to Detect Potential Anomalous Regions Using Context Encoders |
|
Muhammad Zaigham Zaheer, Marcella Astrid(University of Science and Technology, Korea), Seung-Ik Lee(Electronics and Telecommunication Research Institute (ETRI) / University of Science and Technology, Korea), Ho Chul Shin(Electronics and Telecommunication Research Institute (ETRI), Korea) |
| This paper aims to investigate a novel solution for general anomaly detection in surveillance videos by modeling patterns and objects that appear normally in the videos and then using this model to detect the anomalous objects by exploiting image reconstruction methodologies. This approach is inspired by the recently introduced, Context Encoders which are used for semantic image inpainting. Our proposed methodology is semi-supervised which means it does not require an annotated dataset however the videos of cases containing normal scenes are required separately to train the system. |
|