Naslov
The Automatic Image Analysis of Experimental Caustics

Autori
Semenski, Damir ; Magdalenić, Hrvoje

Vrsta, podvrsta i kategorija rada
Radovi u zbornicima skupova, cjeloviti rad (in extenso), znanstveni

Izvornik
Proc. of 14. Danubia-Adria Symposium on Experimental Methods in Solid Mechanics / Biondić, Jasminka - Zagreb : Hrvatsko društvo za mehaniku (HDM), 1997, 34-34

Skup
Danubia-Adria Symposium on Experimental Methods in Solid Mechanics

Mjesto i datum
Poreč, Hrvatska, 02.10.1997. - 04.10.1997

Vrsta sudjelovanja
Poster

Vrsta recenzije
Međunarodna recenzija

Ključne riječi
method of caustics; automated image analysis; stress intensity factor; fracture mechanics

Sažetak
The problem of automated analysis of the experimental effect leads to the image processing of digitized signal by using the appropriate computer hardware and software. The preprogrammed standard functions are not usually derived for the specific problems of each of the optical method, e.g. here the optical method of caustics. Sometimes, the procedure for the image analysis has to be the subject of research, and on the basis of the performed experiments, the standard procedure must be defined. The principal problem of the image analysis in the method of caustics is the problem of detecting the boundary between the dark spot and the surrounding concentrated light. The physical meaning of this boundary is the caustic curve, which shape characterizes the loading and stress conditions of the stress and strain singularity. Sometimes, the shape of this singular curve that can be obtained analytically as the exact mathematical function, or can be simulated by using the numerical procedure of solving the differential equations. The size of the experimental singular curve is of significance for the singularity, while its shape has to be compared to the simulated one, calculated for the unit value of the characterization factor. When the shapes of simulated and experimental singular caustic curve are of the similar shape, the evaluation can be properly carried out. There are three methods for detecting the experimental caustic curves that are introduced and considered in detail: the boundary value method, the percentage method and the differential method. All these methods are based on different algorithms of light intensity determination. The most sophisticated is the differential method, shifting the real image into the gradient image, and collecting the experimental points that are located on the position of the maximum light intensity gradient. The performed experiments improve differential method, as well as the computer programs added to the system ; it is appropriate to be automatic and standard procedure effectively increasing the possibilities of image analysis of experimental caustics.

Izvorni jezik
Engleski

Znanstvena područja
Strojarstvo

POVEZANOST RADA