• Journal of KIISE
• /
• v.42 no.4
• /
• pp.483-486
• /
• 2015

Security systems that use face recognition are vulnerable to spoofing attacks where unauthorized individuals use a photo or video of authorized users. In this work, we propose a method to detect a face spoofing attack with a video of an authorized person. The proposed method uses three sequential frames in the video to extract features by using Fourier Transform and Dense-SIFT filter. Then, classification is completed with a Support Vector Machine (SVM). Experimental results with a database of 200 valid and 200 spoof video clips showed 99% detection accuracy. The proposed method uses simplified features that require fewer memory and computational overhead while showing a high spoofing detection accuracy.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.20 no.9
• /
• pp.49-55
• /
• 2019

This paper presents a marker image evaluation method based on analysis of object distribution in images and classification of images with repetitive patterns for effective marker-based augmented reality (AR) system development. We measure the variance of feature point coordinates to distinguish marker images that are vulnerable to occlusion, since object distribution affects object tracking performance according to partial occlusion in the images. Moreover, we propose a method to classify images suitable for object recognition and tracking based on the fact that the distributions of descriptor vectors among general images and repetitive-pattern images are significantly different. Comprehensive experiments for marker images confirm that the proposed marker image evaluation method distinguishes images vulnerable to occlusion and repetitive-pattern images very well. Furthermore, we suggest that scale-invariant feature transform (SIFT) is superior to speeded up robust features (SURF) in terms of object tracking in marker images. The proposed method provides users with suitability information for various images, and it helps AR systems to be realized more effectively.

• Journal of Internet Computing and Services
• /
• v.20 no.4
• /
• pp.91-102
• /
• 2019

In autonomous driving systems, the ability to classify pedestrians in images captured by cameras is very important for pedestrian safety. In the past, after extracting features of pedestrians with HOG(Histogram of Oriented Gradients) or SIFT(Scale-Invariant Feature Transform), people classified them using SVM(Support Vector Machine). However, extracting pedestrian characteristics in such a handcrafted manner has many limitations. Therefore, this paper proposes a method to classify pedestrians reliably and effectively using CNN's(Convolutional Neural Network) deep features and transfer learning. We have experimented with both the fixed feature extractor and the fine-tuning methods, which are two representative transfer learning techniques. Particularly, in the fine-tuning method, we have added a new scheme, called M-Fine(Modified Fine-tuning), which divideslayers into transferred parts and non-transferred parts in three different sizes, and adjusts weights only for layers belonging to non-transferred parts. Experiments on INRIA Person data set with five CNN models(VGGNet, DenseNet, Inception V3, Xception, and MobileNet) showed that CNN's deep features perform better than handcrafted features such as HOG and SIFT, and that the accuracy of Xception (threshold = 0.5) isthe highest at 99.61%. MobileNet, which achieved similar performance to Xception and learned 80% fewer parameters, was the best in terms of efficiency. Among the three transfer learning schemes tested above, the performance of the fine-tuning method was the best. The performance of the M-Fine method was comparable to or slightly lower than that of the fine-tuningmethod, but higher than that of the fixed feature extractor method.

• Journal of KIISE:Computing Practices and Letters
• /
• v.16 no.9
• /
• pp.889-897
• /
• 2010

Recently, a cutting-edge large-scale image database system has demanded these attributes: search with alarming speed, performs with high accuracy, archives efficiently and much more. An image identifier (descriptor) is for measuring the similarity of two images which plays an important role in this system. The extraction method of an image identifier can be roughly classified into two methods: a local and global method. In this paper, the proposed image identifier, LFH(Local Feature's Histogram), is obtained by a histogram of robust and distinctive local descriptors (features) constrained by a district sub-division of a local region. Furthermore, LFH has not only the properties of a local and global descriptor, but also can perform calculations at a magnificent clip to determine distance with pinpoint accuracy. Additionally, we suggested a way to extract LFH via GPU (OpenGL and GLSL). In this experiment, we have compared the LFH with SIFT (local method) and EHD (global method) via storage capacity, extraction and retrieval time along with accuracy.