• Journal of Computing Science and Engineering
• /
• v.11 no.1
• /
• pp.24-31
• /
• 2017

Face-based video retrieval has become an active and important branch of intelligent video analysis. Face profiling and matching is a fundamental step and is crucial to the effectiveness of video retrieval. Although many algorithms have been developed for processing static face images, their effectiveness in face-based video retrieval is still unknown, simply because videos have different resolutions, faces vary in scale, and different lighting conditions and angles are used. In this paper, we combined content-based and semantic-based image analysis techniques, and systematically evaluated four mainstream local features to represent face images in the video retrieval task: Harris operators, SIFT and SURF descriptors, and eigenfaces. Results of ten independent runs of 10-fold cross-validation on datasets consisting of TED (Technology Entertainment Design) talk videos showed the effectiveness of our approach, where the SIFT descriptors achieved an average F-score of 0.725 in video retrieval and thus were the most effective, while the SURF descriptors were computed in 0.3 seconds per image on average and were the most efficient in most cases.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.34 no.5
• /
• pp.1639-1644
• /
• 2014

SAR satellite can acquire clear imagery regardless of weather and the images are widely used for land management, natural hazard monitoring and many other applications. Automatic image matching technique is necessary for management of a huge amount of SAR data. Nevertheless, it is difficult to assure the accuracy of image matching due to the difference of image-capturing attitude and time. In this paper, we compared performances of MI method, FMT method and SIFT method by applying arbitrary displacement and rotation to TerraSAR-X images and changing resolution of the images. As a result, when the features having specific intensity were distributed well in SAR imagery, MI method could assure 0~2 pixels accuracy even if the images were captured in different geometry. But the accuracy of FMT method was significantly poor for the images having different spatial resolutions and the error was represented by tens or hundreds pixels. Moreover, the ratio of corresponding matching points for SIFT method was only 0~17% and it was difficult for SIFT method to apply to SAR images captured in different geometry.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.17 no.4
• /
• pp.1066-1079
• /
• 2023

Multi-target tracking based on the detector is a very hot and important research topic in target tracking. It mainly includes two closely related processes, namely target detection and target tracking. Where target detection is responsible for detecting the exact position of the target, while target tracking monitors the temporal and spatial changes of the target. With the improvement of the detector, the tracking performance has reached a new level. The problem that always exists in the research of target tracking is the problem that occurs again after the target is occluded during tracking. Based on this question, this paper proposes a DeepSORT model based on SIFT features to improve ship tracking. Unlike previous feature extraction networks, SIFT algorithm does not require the characteristics of pre-training learning objectives and can be used in ship tracking quickly. At the same time, we improve and test the matching method of our model to find a balance between tracking accuracy and tracking speed. Experiments show that the model can get more ideal results.

• ETRI Journal
• /
• v.34 no.4
• /
• pp.572-582
• /
• 2012

The discrete Gaussian-Hermite moment (DGHM) is a global feature representation method that can be applied to square images. We propose a modified DGHM (MDGHM) method and an MDGHM-based scale-invariant feature transform (MDGHM-SIFT) descriptor. In the MDGHM, we devise a movable mask to represent the local features of a non-square image. The complete set of non-square image features are then represented by the summation of all MDGHMs. We also propose to apply an accumulated MDGHM using multi-order derivatives to obtain distinguishable feature information in the third stage of the SIFT. Finally, we calculate an MDGHM-based magnitude and an MDGHM-based orientation using the accumulated MDGHM. We carry out experiments using the proposed method with six kinds of deformations. The results show that the proposed method can be applied to non-square images without any image truncation and that it significantly outperforms the matching accuracy of other SIFT algorithms.

• Journal of information and communication convergence engineering
• /
• v.11 no.4
• /
• pp.298-306
• /
• 2013

The scale invariant feature transform (SIFT) is an effective algorithm used in object recognition, panorama stitching, and image matching. However, due to its complexity, real-time processing is difficult to achieve with current software approaches. The increasing availability of parallel computers makes parallelizing these tasks an attractive approach. This paper proposes a novel parallel approach for SIFT algorithm implementation using a block filtering technique in a Gaussian convolution process on the SIMD Pixel Processor. This implementation fully exposes the available parallelism of the SIFT algorithm process and exploits the processing and input/output capabilities of the processor, which results in a system that can perform real-time image and video compression. We apply this implementation to images and measure the effectiveness of such an approach. Experimental simulation results indicate that the proposed method is capable of real-time applications, and the result of our parallel approach is outstanding in terms of the processing performance.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.2 no.3
• /
• pp.39-52
• /
• 2009

Scale Invariant Feature Transform (SIFT) is an effective algorithm in object recognition, panorama stitching, and image matching, however, due to its complexity, real time processing is difficult to achieve with software approaches. This paper proposes using a reconfigurable hardware processor with integer half kernel. The integer half kernel Gaussian reduces the Gaussian pyramid complexity in about half [] and the reconfigurable processor carries out a parallel implementation of a full search Fast SIFT algorithm. We use a low memory, fine grain single instruction stream multiple data stream (SIMD) pixel processor that is currently being developed. This implementation fully exposes the available parallelism of the SIFT algorithm process and exploits the processing and I/O capabilities of the processor which results in a system that can perform real time image and video compression. We apply this novel implementation to images and measure the effectiveness. Experimental simulation results indicate that the proposed implementation is capable of real time applications.

• Journal of Korea Multimedia Society
• /
• v.18 no.1
• /
• pp.1-8
• /
• 2015

In recent years, ear has emerged as a new biometric trait, because it has advantage of higher user acceptance than fingerprint and can be captured at remote distance in an indoor or outdoor environment. This paper proposes an individual identification method using ear region based on SIFT(shift invariant feature transform). Unlike most of the previous studies using rectangle shape for extracting a region of interest(ROI), this study sets an ROI as a flexible expanded region including ear. It also presents an effective extraction and matching method for SIFT keypoints. Experiments for evaluating the performance of the proposed method were performed on IITD public database. It showed correct identification rate of 98.89%, and it showed 98.44% with a deformed dataset of 20% occlusion. These results show that the proposed method is effective in ear recognition and robust to occlusion.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.24 no.5
• /
• pp.409-416
• /
• 2006

To use image data obtained from different sensors and different techniques, the preprocessing step that registers them in a common coordinate system is needed. For this purpose, we developed the methodology to register middle- and low-resolution satellite images automatically. Firstly, candidate matching points were extracted using the Harris and Harris-affine algorithm. Secondly, we used the correlation coefficient, normalized correlation coefficient and SIFT algorithm to detect conjugate matching points from candidates. Then, to test the feasibility of approaches, we applied the developed methodology to various kinds of satellite images and compared results. The results clearly demonstrate that the methology using the SIFT is appropriate to register these multi-resolution satellite images automatically, compared with the classical cross-correlation.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.11 no.7
• /
• pp.685-692
• /
• 2016

In this paper, we not only take advantage of the SIFT features in banknote recognition, which has robustness to illumination changes, geometric rotation as well as scale changes, but also propose the hierarchical banknote recognition algorithm, which comprised of feature vector extraction from the frame grabbed image of the banknotes, and matching to the prepared data base of multinational banknotes by ANN algorithm. The images of banknote under the developed UV, IR and white illumination are used so as to extract the SIFT features peculiar to each banknotes. These SIFT features are used in recognition of the nationality as well as face value. We confirmed successful function of the proposed algorithm by applying the proposed algorithm to the banknotes of Korean and USD as well as EURO.

• Journal of KIISE
• /
• v.42 no.1
• /
• pp.93-96
• /
• 2015

The SIFT method is well-known for robustness against various image transformations, and is widely used for image retrieval and matching. The SIFT method extracts keypoints using scale space analysis, which is different from conventional keypoint detection methods that depend only on the image space. The SIFT method has also been extended to use higher-order scale space derivatives for increasing the number of keypoints detected. Such detection of additional keypoints detected was shown to provide performance gain in image retrieval experiments. Herein, a sigma based normalization method for keypoint detection is introduced using higher-order scale space derivatives.