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 - In real-world application of affective brain–computer interface (aBCI), individual differences across subjects and non-stationary characteristics of... 相似文献
A twin-plane re-entrant corner effect (TPRE) in growth of chemical vapour deposited (CVD) -SiC is described by the film and particles of gas-phase homogeneous nucleation. The structural morphology has been characterized by scanning electron microscopy and transmission electron microscopy. Morphological characteristics of the deposited crystals, such as triangularity, hexagons or facets have been explained in terms of the re-entrant corner effect at twin junctions, which were proposed as preferential growth sites for perfect crystals. For real deposits, screw dislocations and/or the re-entrant corner effect are not expected to be compatible. The majority of chemical vapour deposited SiC crystals have a high defect density comprised of {111} twins and dislocations associated with the process variables. Infrared transmission spectra and electron spectroscopy of chemical analysis indicated that the major chemical bonds of CVD -SiC were Si-C and C-H bonds. The positions of the 1s or 2p corelevel peaks for deposits are described. 相似文献
The bulk elastic and piezoelectric properties of potassium titanyl phosphate (KTP) and some of its family members such as RbTiOPO(4 ) (RTP) and RbTiOAsO(4) (RTA) have been determined. The piezoelectric, elastic, and dielectric matrix (P-E-D) of KTP has been completely characterized. Values for the temperature dependence of the expansion coefficients and elastic constants have been obtained, and the effects of variations in dielectric constant on coupling coefficient determined. The major diagonal elements of the piezoelectric and elastic matrices of RTP and RTA were measured as were the thermal expansion coefficient of all the KTP isomorphs, KTP in the temperature range 25-80 degrees C, is shown to have a combination of large coupling coefficients (up to k(t)=38% for hydrothermally grown KTP) and medium temperature dependence coefficients of elastic constant (as low as 40 p.p.m./ degrees C for tau(c33) for hydrothermally grown KTP) that makes it attractive for many piezoelectric applications. 相似文献