CCD microscopy - Image analysis by group for intelligent systems (GIS)

Abstract

BACKGROUND: Classic microscopy does not allow dealing with visual and photographic phenomena, observation of the distinctions where they are needed, or carrying out of any image processing other than cutting. Digital imaging provided more detailed and distinctive insights, and formed the basis for the development of the image processing tools and techniques that transcended in potential and power all initial expectations. The purpose of this article is to present our achievements in this area. METHODS: The starting point is that the matrix. Digital image is a discrete approximation of the continuous two-argument function. Analogue original is a fair basis to start building good mathematical representations of present objects, structures, features, which are then subjected to calculations, transformations and analyses that could precisely match the predefined aims. These analyses are: photomorphology 3D model with morphometrics and full 3D navigation, mathematical representations of granular forms in images, object contours, mathematical representations of chromosomes with mathematical definition of similarity, automatic procedures, such as pattern normalizations, matching, classifications, which leads to broader application of Artificial Intelligence methodology. RESULTS: We developed a method and a complex software environment for microscopic imaging, with many tools and algorithms that proved to be useful in genetics, pathology, and oncology. The presented method is prepared and available for further generalizations and automatization, easily bridging to intelligent systems. CONCLUSION: Microscopic imaging is powerful new high-tech domain of great assistance in biomedical research and medical practice that is revolutionizing real time diagnostic methods and potential, matching the power of molecular biology techniques. Being the pioneers in the microscopic imaging, we are pleased that it is exponentially expanding to the general benefit.