• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제8권2호
• /
• pp.483-503
• /
• 2014

In this paper, human action recognition using pyramid histograms of oriented gradients and collaborative multi-task learning is proposed. First, we accumulate global activities and construct motion history image (MHI) for both RGB and depth channels respectively to encode the dynamics of one action in different modalities, and then different action descriptors are extracted from depth and RGB MHI to represent global textual and structural characteristics of these actions. Specially, average value in hierarchical block, GIST and pyramid histograms of oriented gradients descriptors are employed to represent human motion. To demonstrate the superiority of the proposed method, we evaluate them by KNN, SVM with linear and RBF kernels, SRC and CRC models on DHA dataset, the well-known dataset for human action recognition. Large scale experimental results show our descriptors are robust, stable and efficient, and outperform the state-of-the-art methods. In addition, we investigate the performance of our descriptors further by combining these descriptors on DHA dataset, and observe that the performances of combined descriptors are much better than just using only sole descriptor. With multimodal features, we also propose a collaborative multi-task learning method for model learning and inference based on transfer learning theory. The main contributions lie in four aspects: 1) the proposed encoding the scheme can filter the stationary part of human body and reduce noise interference; 2) different kind of features and models are assessed, and the neighbor gradients information and pyramid layers are very helpful for representing these actions; 3) The proposed model can fuse the features from different modalities regardless of the sensor types, the ranges of the value, and the dimensions of different features; 4) The latent common knowledge among different modalities can be discovered by transfer learning to boost the performance.

• KOREAN JOURNAL OF CROP SCIENCE
• /
• 제50권1호
• /
• pp.60-66
• /
• 2005

Pod shape of twenty soybean (Glycine max L. Merrill) genotypes was evaluated quantitatively by image analysis using elliptic Fourier descriptors and their principal components. The closed contour of each pod projection was extracted, and 80 elliptic Fourier coefficients were calculated for each contour. The Fourier coefficients were standardized so that they were invariant of size, rotation, shift, and chain code starting point. Then, the principal components on the standardized Fourier coefficients were evaluated. The cumulative contribution at the fifth principal component was higher than $95\%$, indicating that the first, second, third, fourth, and fifth principal components represented the aspect ratio of the pod, the location of the pod centroid, the sharpness of the two pod tips and the roundness of the base in the pod contour, respectively. Analysis of variance revealed significant genotypic differences in these principal components and seed number per pod. As the principal components for pod shape varied continuously, pod shape might be controlled by polygenes. It was concluded that principal component scores based on elliptic Fourier descriptors yield seemed to be useful in quantitative parameters not only for evaluating soybean pod shape in a soybean breeding program but also for describing pod shape for evaluating soybean germplasm.

• Proceedings of the Korean Society of Tribologists and Lubrication Engineers Conference
• /
• 한국윤활학회 2002년도 proceedings of the second asia international conference on tribology
• /
• pp.257-258
• /
• 2002

The surface morphology of oil-lubricated surface for hydraulic piston motor is believed to be extremely effective in contact mechanics. adhesion. friction and wear. In order to describe morphology of various rubbed surface on driving condition, the wear test was carried out under different experimental conditions in oil-lubricated system. And fractal descriptors was applied to rubbed surface of hydraulic members with image processing system. These descriptors to analyze surface structure are fractal dimension. Surface fractal dimension can be determined by sum of intensity difference of surface pixel. Morphology of rubbed surface can be effectively obtained by fractal dimension.

• KSTLE International Journal
• /
• 제3권2호
• /
• pp.75-78
• /
• 2002

The surface morphology of oil-lubricated surfaces fer hydraulic piston motors is believed to be extremely effective in contact mechanics, adhesion, friction and weak In order to describe morphology of various rubbed surfaces on driving conditions, the wear test was carried out under different experimental conditions in an oil-lubricated system. And fractal descriptors were applied to rubbed surfaces of hydraulic members and analyzed through an image processing system. These descriptors to analyze surface structure are fractal dimension. Surface fractal dimensions can be determined by sum of intensity difference of surface pixel. The morphology of rubbed surfaces can be effectively obtained by fractal dimension.

• Proceedings of the IEEK Conference
• /
• 대한전자공학회 2003년도 신호처리소사이어티 추계학술대회 논문집
• /
• pp.437-440
• /
• 2003

When we search database with a query image, the retrieval efficiency will vary from each kind of descriptor. Even the best representative descriptor, it results a few useless images that don't match with query image. This type of error can be reduced by adopting another descriptor which extracts features in different way. At present, the choice of descriptors is base on intuitive and experimental method. By theoretic accessing to the problem of descriptor choice, we can solve the given problem in the objective and rational way. In this study, we intend to make a composite of descriptors that can reduce retrieval error by adopting principal component analysis.

• Proceedings of the Korean Society of Tribologists and Lubrication Engineers Conference
• /
• 한국윤활학회 1998년도 제28회 추계학술대회
• /
• pp.115-123
• /
• 1998

The morphological analysis of wear particle is a very effective means for machine condition monitoring and fault diagnosis. In order to describe morphology of various wear particle, the wear test was carried out under different experimental conditions. And fractal descriptors was applied to boundary and surface of wear particle with image processing system. These descriptors to analyze shape and surface wear particle are shape fractal dimension and surface fractal dimension. The shape fractal dimension can be derived from the boundary profile and surface fractal dimension can be determined by sum of intensity difference of surface pixel. The morphology of wear particles can be effectively obtained by two fractal dimensions.

• Proceedings of the KSME Conference
• /
• 대한기계학회 2000년도 춘계학술대회논문집A
• /
• pp.755-759
• /
• 2000

The determination of surface topography is believed to be extremely important in the areas of contact mechanics, adhesion and friction. In order to describe topography of various frictional surface. the wear test was carried out under different experimental conditions in dry friction. And fractal descriptors was applied to frictional surface of laser modified steel with image processing System. These descriptors to analyze surface structure are fractal dimension. Surface fractal dimension can be determined by sum of intensity difference of surface pixel. Topography of frictional surface can be effectively obtained by fractal dimensions.

• Proceedings of the Korean Society of Tribologists and Lubrication Engineers Conference
• /
• 한국윤활학회 1999년도 제30회 추계학술대회
• /
• pp.240-246
• /
• 1999

The determination of surface topography is believed to be extremely important in the areas of contact mechanics, adhesion and friction. In order to describe topography of various frictional surface, the wear test was carried out under different experimental conditions in dry friction. And fractal descriptors was applied to frictional surface of laser modified steel with image processing system. These descriptors to analyze surface structure are fractal dimension. Surface fractal dimension can be determined by sum of intensity difference of surface pixel. Topography of frictional surface can be effectively obtained by fractal dimensions.

• Proceedings of the Korean Society of Machine Tool Engineers Conference
• /
• 한국공작기계학회 2000년도 추계학술대회논문집 - 한국공작기계학회
• /
• pp.404-409
• /
• 2000

The determination of surface morphology is believed to be extremely important in the areas of contact mechanics, adhesion and friction. In order to describe morphology of various rubbed surface, the wear test was carried out under different experimental conditions in lubricating wear. And fractal descriptors was applied to rubbed surface of hydraulic driving material with image processing system. These descriptors to analyze surface structure are fractal dimension. Surface fractal dimension can be determined by sum of intensity difference of surface pixel. Morphology of rubbed surface can be effectively obtained by fractal dimensions.

• Tribology and Lubricants
• /
• 제18권2호
• /
• pp.147-152
• /
• 2002

The morphological analysis of wear particle is a very effective means fur machine condition monitoring and fault diagnosis. In order to describe morphology of various wear particle, the wear test was carried out under friction experimental conditions. And fractal descriptors was applied to boundary and surface of wear particle with image processing. These descriptors to analyze shape and surface of wear particle are shape fractal dimension and surface fractal dimension. The boundary fractal dimension can be derived from the boundary profile and surface fractal dimension can be determined by sum of intensity difference of surface pixel. The morphology of wear particles can be effectively obtained by two fractal parameter.