AUTOMATIC TEXT LOCATION IN IMAGES AND VIDEO FRAMES

Textual data is very important in a number of applications such as image database indexing and document understanding. The goal of automatic text location without character recognition capabilities is to extract image regions that contain only text. These regions can then be either fed to an optical character recognition module or highlighted for a user. Text location is a very difficult problem because the characters in text can vary in font, size, spacing, alignment, orientation, color and texture. Further, characters are often embedded in a complex background in the image. We propose a new text location algorithm that is suitable in a number of applications, including conversion of newspaper advertisements from paper documents to their electronic versions, World Wide Web search, color image indexing and video indexing. In many of these applications, it is not necessary to extract all the text, so we emphasize on extracting important text with large size and high contrast. Our algorithm is very fast and has been shown to be successful in extracting important text in a large number of test images.     

Electrical-Mechanical Analog Applied to Crack Growth in Glass- Glass-Ceramic Composites

It has been shown that the electrical-mechanical analog is exact under certain conditions. This exactness was confirmed by determining voltage as a function of crack size on a center-cracked and an edge-cracked aluminum foil. The technique was also used to determine the crack velocity-stress intensity factor diagram from the growth of a microcrack in a glass-glass-ceramic composite system. The crack length-time relation was determined by measuring voltage-time across a vapor-deposited gold film on the specimen surfaces.     

Crack-Propagation Rates in β"-Alurnina

Specimens of sodium and silver β"-alumina were electrolyLed for 3ariom times and subsequently broken in bending. Using the Griffith equation, the lengths of the cracks after electrolysis were estimated as a function of electroljsis time. It way obsened that the crack length vs time plots were bimodal. During the early Ytages of crack propagation, the crack velocity war found to he independent of crack length. The rate of crack propagation in Na-β"-alumina was about three times batter than that in Ag-β"-alumina. Furthermore, in Na-β"-alumina, the rate of crack propagation was found to increase with increasing temperature. The dependence of crack velocity on temperature, the difference in crack-growth rates in Na-β-alumina and Ag-β"-alumina, and the fact that crack velocity is independent of crack length were rationalized in terms of a model that takes into account the transient nature of ionic flux to the growing crack.     

Investigation of Phase Stability in the System Sic-AlN

Dense samples of several compositions in the system SiC-AIN were fabricated by hot-pressing. The SiC-AIN powder was prepared by carbothermal reduction of an intimate mixture of alumina, silica, and carbon in a nitrogen atmosphere. X-ray diffraction and electron and optical microscopy were used to determine the chemical and microstructural characteristics of the hot-pressed specimens. Materials with bulk compositions between 15 and 75 wt% AIN were found to be nonhomo-geneous when hot-pressed below 2100°C. These materials were determined to be a mixture of SiC-AIN solid solutions with different compositions. The observed compositional variations were distinctly bimodal. The source of the in-homogeneity was the starting SiC-AIN powder. The powders, as well as the hot-pressed samples, consisted of a mixture of small crystals rich in SiC and large AIN-rich crystals. Compositions outside the 15 to 75 wt% AIN region were found to be single phase and to have the wurtzite structure. Hot-pressing SiC-AIN in the intermediate composition range at 2300°C produced an optically and chemically homogeneous material. The precipitation of an SiC-rich phase from a 75 wt% AIN solid solution and the precipitation of an AIN-rich phase from a 47 wt% AIN alloy when annealed at 1700°C are strong indications that a miscibility gas exists in the system SiC-AIN.     

Application of Load-Relaxation Techniques to Study Subcritical Crack Growth in Brittle Materials

The load relaxation technique was applied to double- cantilever-beam specimens for the determination of diagrams of the stress intensity factor (K₁ ) vs crack velocity (V) for soda-lime glass. The effect of machine compliance was explicitly incorporated into the velocity equation. The data were in good agreement with published data from different techniques. It was shown that a hard machine (small compliance) is best suited for the load-relaxation technique. The use of a soft machine (high compliance) will lead to catastrophic failure during load relaxation. A single-edge notched-beam specimen in bending is unsuitable for the relaxation technique since catastrophic failure will occur during relaxation regardless of machine compliance.     

Polarization and Conduction Mechanism in Mixed-Alkali Glasses

Thermally stimulated polarization and depolarization current (TSPC/TSDC) measurements on mixed alkali (Li-Na) silicate glasses show two TSDC peaks along with a dc conduction current at high temperatures, similar to single-alkali silicate glasses. The activation energies for the two TSDC peaks and for dc conduction increase with the addition of a second alkali. The first TSDC peak does not change in size but the second becomes substantially smaller when the second alkali is added. The reduction in size of the second TSDC peak correlates well with the reduction in the dielectric loss hi mixed-alkali glasses. A model is proposed to account for the behavior of the two TSDC peaks.     

AI-based adaptive control and design of autopilot system for nonlinear UAV

The objective of this paper is to design an autopilot system for unmanned aerial vehicle (UAV) to control the speed and altitude using electronic throttle control system (ETCS) and elevator, respectively. A DC servo motor is used for designing of ETCS to control the throttle position for appropriate amount of air mass flow. Artificial Intelligence (AI)-based controllers such as fuzzy logic PD, fuzzy logic PD + I, self-tuning fuzzy logic PID (STF-PID) controller and fuzzy logic-based sliding mode adaptive controller (FLSMAC) are designed for stable autopilot system and are compared with conventional PI controller. The target of throttle, speed and altitude controls are to achieve a wide range of air speed, improved energy efficiency and fuel economy with reduced pollutant emission. The energy efficiency using specific energy rate per velocity of UAV is also presented in this paper.