The rendering of the project or a portion of it can be exported in several image formats or as vector graphics in the SVG format. In JMicroVision, characteristics of the objects extracted from the image can be highlighted in a visual way either by various diagrams or by a coloring of objects according to their value (in relation with a selected descriptor). It offers several advantages compared to the counting under the microscope, like the restriction to a region type (after segmentation) in order to quickly estimate its subclasses. The latter represents an alternative when the various segmentation methods do not allow a satisfying estimate of the modal composition. Other innovating tools, such as granulometric analysis and computerized point counting, were developed. In order to optimize memory management during certain operations, we adapted the algorithm of region growing so that it is compatible with tiling and deferred execution. This thesis presents a new method to separate connected grains, which is more efficient than traditional methods on long particles and on those containing holes. Even with very large images and a computer with little memory, the visualization tools remain powerful thanks to tiling and deferred execution. Moreover, the magnifying lens and the multiview tool allow a simultaneous view of several images, each one having its own zoom coefficient while keeping a joint position in the center. Like a microscope, it enables a dynamic observation of a thin section with the possibility of combining various focus or modes of lighting (polarized light, cathodoluminescence…). JMicroVision was designed to display and process high-resolution images under optimal conditions. These tools were developed in the Java programming language and were gathered in the JMicroVision software. This thesis tries to fill this gap by providing a suite of computer tools more adapted to the visualization and the analysis of petrographic images.
#Jmicrovision website software#
However, because of the complexity and the large variety of images, existing software is not always very efficient. One could expect that petrographic components can easily be quantified and characterized by a computer program. In the field of petrography, microscopic observation remains fundamental for determining rocks and understanding their genesis.
0 kommentar(er)
