To improve the accuracy of retinal vessel segmentation, a retinal vessel segmentation algorithm for color fundus images based on back-propagation (BP) neural network is proposed according to the characteristics of retinal blood vessels. Four kinds of green channel image enhancement results of adaptive histogram equalization, morphological processing, Gaussian matched filtering, and Hessian matrix filtering are used to form feature vectors. The BP neural network is input to segment blood vessels. Experiments on the color fundus image libraries DRIVE and STARE show that this algorithm can obtain complete retinal blood vessel segmentation as well as connected vessel stems and terminals. When segmenting most small blood vessels, the average accuracy on the DRIVE library reaches 0.9477, and the average accuracy on the STARE library reaches 0.9498, which has a good segmentation effect. Through verification, the algorithm is feasible and effective for blood vessel segmentation of color fundus images and can detect more capillaries.
Neural Computing and Applications - RNA-binding proteins play an important role in the biological process. However, the traditional experiment technology to predict RNA-binding residues is... 相似文献
Non-homogeneity detector (NHD) is known for improving the detection performance of space time adaptive processing (STAP) in heterogeneous environment. To mitigate the finite sample effect in the generalized inner product (GIP) based NHD, a new type of GIP detector based on diagonal loading (LGIP) is proposed in this paper. We derive the theoretical mean value of LGIP detector based on random matrix theory. Moreover, an approximation of the theoretical mean value is provided for practical considerations. We then apply the derived theoretical mean value of LGIP to the detection of heterogeneous samples. Finally, by theoretical analysis and numerical simulations, we show the effectiveness of the proposed detector. 相似文献
The thermal conductivity coefficient for a Kondo superconductor in the gapless state is obtained within the framework of the Matsuura, Ichinose, and Nagaoka theory. By expanding the heat current correlation function appearing in the Kubo formula for the thermal conductivity in powers of the order parameter, analytical expressions for the thermal conductivity to second power in the order parameter are obtained. The thermal conductivities for bothTKTc0 andTKTc0 superconductors are obtained. 相似文献
Experimental investigations have been made in which single crystals of solid explosives such as cyclotrimethylene trinitramine, RDX, pentaerythritol tetranitrate, PETN, AgN3 and Pb(N3)2, were irradiated with a focused laser beam of wavelength 266 nm and duration 5 nsec in an ultra high vacuum system housing a fast and sensitive time-of-flight mass spectrometer. From the mass spectra (both +ve and -ve ions) of the chemical species obtained, possible decomposition schemes of RDX and PETN have been proposed. For silver and lead azides, the laser irradiation did not cause explosions if the power density was sufficiently low, but micrometre size craters were formed. In the mass spectra of AgN3 products N3– and silver cluster ions, Agn+, have been found for the first time. It has been suggested that such investigations may possibly throw some light on our understanding of the reactions taking place in a deflagration/detonation wave. 相似文献