Image-level structural recognition is an important problem for many applications of computer vision such as autonomous vehicle control, scene understanding, and 3D TV. A novel method, using image features extracted by exploiting predefined templates, each associated with individual classifier, is proposed. The template that reflects the symmetric structure consisting of a number of components represents a stage—a rough structure of an image geometry. The following image features are used: a histogram of oriented gradient (HOG) features showing the overall object shape, colors representing scene information, the parameters of the Weibull distribution features, reflecting relations between image statistics and scene structure, and local binary pattern (LBP) and entropy (E) values representing texture and scene depth information. Each of the individual classifiers learns a discriminative model and their outcomes are fused together using sum rule for recognizing the global structure of an image. The proposed method achieves an 86.25% recognition accuracy on the stage dataset and a 92.58% recognition rate on the 15-scene dataset, both of which are significantly higher than the other state-of-the-art methods.
1 August 2014 in Journal of Experimental and Theoretical Physics
It has been shown that the conclusion of the linear instability of the Hagen-Poiseuille flow at finite Reynolds numbers requires the refusal of the use of the traditional “normal” form of the representation of disturbances, which implies the possibility of separation of variables describing disturbances as functions of the radial and longitudinal (along the axis of a tube) coordinates. In the absence of such separation of variables in the developed linear theory, it has been proposed to use a modification of the Bubnov-Galerkin theory that makes it possible to take into account the difference between the periods of the longitudinal variability for different radial modes preliminarily determined by the standard application of the Galerkin-Kantorovich method to the evolution equation of extremely small axisymmetric disturbances of the tangential component of the velocity field. It has been shown that the consideration of even two linearly interacting radial modes for the Hagen-Poiseuille flow can provide linear instability only in the presence of the mentioned conditionally periodic longitudinal variability of disturbances along the axis of the tube, when the threshold Reynolds number Reth(p) is very sensitive to the ratio p of two longitudinal periods each describing longitudinal variability for its radial disturbance mode. In this case, the threshold Reynolds number can tend to infinity, Reth(p) → ∞, only at p = pk = k, p = p1/k = 1/k, and \(p = p_{\sqrt k } = [k + 1 \pm \sqrt {(k + 1)^2 - 4} ]/2\) , where k = 1, 2, 3, …. The minimum Reynolds number Reth(p) ≈ 448 (at which p ≈ 1.527) for the linear instability of the Hagen-Poiseuille flow quantitatively corresponds to the condition of the excitation of Tollmien-Schlichting waves in the boundary layer, where Reth = 420. Similarity of the mechanisms of linear viscous dissipative instability for the Hagen-Poiseuille flow and Tollmien-Schlichting waves has been discussed. Good quantitative agreement has been obtained between the phase velocities of the vortex disturbances and the experimental data on the velocities of the leading and trailing edges of turbulent “puffs” propagating along the axis of the tube.
Improved fast encryption algorithm for multimedia IFEA-M is modified to withstand recently discovered vulnerability to differential known plaintext-ciphertext attack. The proposed parameterized algorithm PIFEA-M has the same performance as FEA-M that is about 25% better than that of IFEA-M.
A new pseudo-random number generator algorithm RC4E, having period significantly greater than RC4 has, is proposed. It is an extension of RC4. It uses RC4 as a main engine and Heap's algorithm as a second engine, providing enumeration of all possible permutations. Second engine is invoked periodically and it resets the current state of the main engine. Experiments, conducted on the both algorithms, showed their good statistical properties and practically same performance
A general exact weak solution to the non-linear equation of the conservation of the absolute vorticity in a thin layer of an incompressible medium on a rotating sphere is proposed. It takes into account the helicity of the point vortices and the non-uniformity of the depth of the layer. This is used to develop a model of the observed interactions of spiral atmospheric vortices. The fusion of two-point helical vortices (HVs) on the rotating sphere is considered. We also analyze the prognostic applicability of the dynamics of the HVs for modeling the abrupt changes observed in the trajectories of tropical cyclones and their landfall in comparison with the traditional approach. The analytical condition for chiral symmetry violation related to the direction of the movement of the center of a helical cyclone is obtained.