Pavel Starha

Application of image analysis method to detection and counting of glass fibers from filter samples

ABSTRACT Man-made vitreous fibers (MMVFs) are noncrystalline substances made of glass, rock or slag and are widely used as thermal or acoustic insulation materials. There is continued concern about their potential health impacts and thus, their dosimetry and behavior in the environment still require study using filters to collect fiber samples. After deposition or exposure measurements of MMVFs it is often necessary to analyze the filters with deposited fibers. This task is tedious, time-consuming, and requires skill. Therefore, many researchers have tried to simplify or automatize fiber detection and quantification. This article describes features of our in-house software, which automatically detects and counts fibers in images of filter samples. The image analysis is based on the use of a histogram equalization and an adaptive radial convolution filter that enhances fiber contrast and thus, improves the fiber identification. The accuracy of the software analysis was verified by comparison with manual counting using ordinary phase-contrast microscopy method. The correlation between the methods was very high (coefficient of determination was 0.977). However, there were some discrepancies caused by false identifications, which led to implementation of manual corrective functions. Copyright

Histometric evaluation of meat products: Determination of size and number of objects

In the framewort of the quantitative histologic evaluation of poultry products, the size and the number of bone fragments have been determined using the image analysis. Bone fragments were identified by their colour and analysed automatically. The samples contained 135 to 2167 bone particles the length of which varied from 5 to 2088 μm. Comparing products of the same kind, we found differences in the contents of bone fragments; this fact was possibly due to inadequate observance of the technological procedure by some producers.

Decomposition of a Bunch of Objects in Digital Images

In object recognition methods, we often come across a task to identify objects that originated from several overlapping elementary objects. After image segmentation, this bunch of elementary objects is identified as one object, which has completely different geometrical properties from the elementary objects. The information about the number of objects, their size and shape is degraded. The paper considers a step towards solving this problem. It describes a mathematical method for decomposition of objects formed by slightly overlapping / touching objects in separate elementary objects. The method is based only on the geometrical properties of the object of interest and on the assumption of the shape of the elementary objects.

Qualitative Analysis of Film Thickness in Rolling EHD Contact by Fluorescence Technique

Nowadays, it is well known that tribological processes play a significant role in a daily life. The elastohydrodynamic (EHD) regime of lubrication occurs in many applications including both technical and biological points of view. During the last fifty years there was a rapid evolution in experimental methods which help to better understand friction, lubrication and wear. In this study, mercury lamp induced fluorescence is employed to qualitative assess the film thickness in rolling EHD contact created by a steel ball and glass disk. Several experiments were performed to confirm the theory that the dependence between film thickness and entrainment speed should be linear in logarithmic. The results show satisfactory agreement with this prediction, so the method is correct and suitable for further research in the area of EHD contacts. This method also allows analyzing of compliant non-reflective contacts, which will be described in the next paper.

Utilization of mathematical processing of retinal images with use of adaptive contrast control (ACC) method to detect quantity of vascular endings

Purpose The authors investigated whether the method of mathematical processing of retinal images with the use of a computer can be used to evaluate ocular background figures of patients with physiologic retinal findings. This method is based on identification of vascular endings in an examined retinal area. When the authors mention vascular endings, they do not refer to factual endings, but recognizable vascular endings; there are no endings in a vascular system. Methods Adaptive contrast control (ACC) method was used to determine a number of vascular endings. The method is based on mathematical processing of a digitized retina picture with the use of a computer. On a digitized retinal picture, the vascular system is identified with the use of the conditional erosion methodology, and the number of vascular endings is then determined. The ACC method was used to process a file of retinal pictures of 38 patients (76 eyes) with physiologic retinal findings. Results Based on the results of statistical analysis, the authors detected that the number of vascular endings showed a normal curve (Gaussian distribution, p=0.05). A tight correlation between quantities of vascular endings in the right and the left eyes was also detected, which means that the quantity of vascular endings in the right and the left eyes is in a very close correlation (p=0.05). Conclusions The authors highlight that the curve of the number of vascular endings of patients with physiologic retinal findings shows a Gaussian distribution.