Active Appearance Model (AAM) is an accurate and robust tool and is suitable when it’s needed to estimate shape of object when its’ approximate shape is known but varies within a certain range from instance to instance. An AAM allows complex models of shape (for example human face) and appearance to be matched to new images rapidly. An AAM contains a statistical model of the shape and gray level or color appearance of an object of interest. The associated search algorithm exploits the locally linear relationship between model parameter displacements and the residual errors between model instance and image. AAM is widely used but the research of its’ accuracy and stability still remains an important and not fully learned issue. In this paper, we study landmarks stability and error estimation produced by AAM in different lightning conditions and signal-to-noise ratio (SNR).
Using of Michelson interferometer is shown in the ï¬eld of measurement of periodical displacements of the con-trolled object. The foundations of optical interferometry are presented. The features of Michelson interferometer are described. The mathematical model of interference pattern produced by Michelson interferometer is created. It takes in consideration such parameters as the angles at which the mirrors are located and the lengths of two optical paths.
The main reasons of catastrophes and accidents are high level of wear of equipment and violation of the production technology. The methods of nondestructive testing are designed to find out defects timely and to prevent break down of aggregates. These methods allow determining compliance of object parameters with technical requirements without destroying it. This work will discuss dye penetrant inspection or liquid penetrant inspection (DPI or LPI) methods and computer model of microcracks image treated with fluorescent dye.
Usually cracks on image look like broken extended lines with small width (about 1 to 10 pixels) and ragged edges. The used method of inspection allows to detect microcracks with depth about 10 or more micrometers.
During the work the mathematical model of image of randomly located microcracks treated with fluorescent dye was created in MATLAB environment. Background noises and distortions introduced by the optical systems are considered in the model.
The factors that have influence on the image are listed below:
1. Background noise. Background noise is caused by the bright light from external sources and it reduces contrast on the objects edges.
2. Noises on the image sensor. Digital noise manifests itself in the form of randomly located points that are differing in their brightness and color.
3. Distortions caused by aberrations of optical system. After passing through the real optical system the homocentricity of the bundle of rays is violated or homocentricity remains but rays intersect at the point that doesn’t coincide with the point of the ideal image.
The stronger the influence of the above-listed factors, the worse the image quality and therefore the analysis of the image for control of the item finds difficulty.
The mathematical model is created using the following algorithm: at the beginning the number of cracks that will be modeled is entered from keyboard. Then the point with random position is choosing on the matrix whose size is 1024x1024 pixels (result image size). This random pixel and two adjacent points are painted with random brightness, the points, located at the edges have lower brightness than the central pixel. The width of the paintbrush is 3 pixels. Further one of the eight possible directions is chosen and the painting continues in this direction. The number of ‘steps’ is also entered at the beginning of the program. This method of cracks simulating is based on theory A.N. Galybin and A.V. Dyskin, which describe cracks propagation as random walk process. These operations are repeated as many times as many cracks it’s necessary to simulate. After that background noises and Gaussian blur (for simulating bad focusing of optical system) are applied.