In case that a pattern recognition system is used for images recognition (in robot vision, handwritten recognition etc.),
the system must have the capacity to identify an object indifferently of its size or position in the image.
The problem of the invariance of recognition can be approached in some fundamental modes. One may apply the
similarity criterion used in associative recall. The original pattern is replaced by a mathematical transform that assures
some invariance (e.g. the value of two-dimensional Fourier transformation is translation invariant, the value of Mellin
transformation is scale invariant). In a different approach the original pattern is represented through a set of features,
each of them being coded indifferently of the position, orientation or position of the pattern.
Generally speaking, it is easy to obtain invariance in relation with one transformation group, but is difficult to obtain
simultaneous invariance at rotation, translation and scale.
In this paper we analyze some methods to achieve invariant recognition of images, particularly for digit images. A great
number of experiments are due and the conclusions are underplayed in the paper.
The objectives of this study are to survey the literature regarding energy evaluation and find
optimal solution based on models and experimental set-up's for small and medium home in Alba area.
Wind resource evaluation is a critical element in projecting turbine performance at a given site.
The energy available in a wind stream is proportional to the cube of its speed, which means that
doubling the wind speed increases the available energy by a factor of eight. Furthermore, the
wind resource itself is seldom a steady, consistent flow. It varies with the time of day, season,
height above ground, and type of terrain. Proper siting in windy locations, away from large
obstructions, enhances a wind turbine's performance. We used our measured data in 2 locations
around our town and with this data inserted in some Matlab models we intend to build optimal
experimental solution for produced home's generator's, around Alba town, up to 5 kW.
This paper presents the new technologies those were applied in manufacturing of lighting systems of road vehicles. Litronic gets an important role in the front lighting of new cars. New perspectives were opened by PES - reflectors and FT technology. It is also very interesting the application of Power LEDs in car lighting systems.
One of the most promising implementation of artificial neural networks is optoelectronic implementation. Optical interconnections are useful for neural networks as far as one can take advantage of the special potential of 3D connection through free space. In hardware implementations of neural networks, the weights values will be materialized in a technological process during which various errors may occur, so that the resulting network will use more or less deviated weights. This paper studies several aspects concerning the optical interconnection of artificial neurons. The authors analyze the problems involved by using computer-generated holograms (CGH) for these interconnections and some methods of designing such diffractive elements. The authors also analyze the error sources and the consequences caused by random deviations of the neurons interconnection weights from the accurately computed values. The theoretical considerations are illustrated by designing an auto associative memory built for graphic pattern recognition. Neurons interconnections are to be implemented optically by computer generated holograms (CGH). The network functioning was simulated on computer and the paper presents also the results of simulations on a data set and a CGH layout for neuron interconnections.