This study analyzes the evaluation of land cover supervised classification quality. Authors put forward the hypothesis that the overall accuracy of image classification depends on its division into parts of the same area. The dependence is described by the logarithmic curve – Т = 4.3004·ln(x) + 72.697, because the determination coefficient is maximum (R2 = 0.9678). The research area was the Yuntolovo reserve, the protected area near St. Petersburg (Russia). In order to increase the overall accuracy of the land cover automatic classification based on aerial images, a new methodology of data preprocessing was introduced. The proposed method of estimating the overall classification accuracy of land cover protected areas increases on average by 10% by dividing the source aerial image into no more than 10 equal parts. With further partitioning of the image into parts of the same area, the overall accuracy is slightly increased. Pixel-based image analysis of supervised classification and error matrix were evaluated using ILWIS 3.31 software and in our own software in .NET environment.
There are many machines, devices and production lines that are equipped with rotating elements (shafts, axes, spindles, trailing and drive wheels, etc.). Correct geometry of these elements ensures trouble-free operation, and, in the case of machine tools it decides about the correct geometric parameters of the manufactured semi-finished articles and products. In this respect, the newly manufactured and operated lathes are checked against the correct geometry of their parts which determine the location of the work-piece and the lathe and their movement These parts are geometrically correct if the errors of their geometrical shape (deviations) do not exceed values designated in the standards. The wearing of lathes reveals the so-called spindle run-out (deviation from circularity). These deviations are determined using mainly the workshop methods. However, due to the considerable sizes of machine tools surveying methods and new methods which use electronic and optoelectronic devices also apply. Authors of this work present the measuring set (which they designed and built themselves) which relies on the two-point fibre signalling device which is installed in the lathe spindle chuck jaws. The position of the signalling device during the lathe spindle rotation is recorded with CCD/CMOS digital camera, and the images are stored on the camera memory card. The aim of the presented research and experimental works was to determine the internal conformity of measurement results that were obtained using the designed and constructed measuring set, to check whether there is a correlation between the results obtained using the dial indicator and to analyse the accuracy of observations that were made using the test set. The designed measuring set enables to determine deviations from circularity during the lathe spindle rotation with the accuracy of ± 0.02 mm.
THE IDENTIFICATION OF PARAMETERS OF A LINEAR AND A NON-LINEAR MODEL OF A KINEMATIC MEASUREMENT-CONTROL NETWORK
kinematic model of a geodetic network, vertical displacements, neural networks
Engineering geodesy deals with a wide range of problems. There is also a part that deals with measuring displacements and deformations of engineering objects. Correct geodetic monitoring requires identifying the movement of points representing an engineering object in order to determine displacement values, taking into account the time function. The paper presents the results of research on kinematic models of geodetic networks in the aspect of using them for describing the state of vertical displacements of engineering objects located on expansive soil. The paper presents two functional models of an observation system: one in the form of a second rank polynomial and the other in the form of an exponential function. The selected kinematic models of measurement-control geodetic networks were estimated with classic methods and neural networks.