Research On Pixel-Oriented Visualization Techniques

A wide variety of advanced visual exploration and visualization methods have been proposed in the past. These techniques have proven to be of high value in supporting researchers and analysts to obtain insight into large data sets and to turn raw data into useful and valuable knowledge.We first provide an overview of current visualization methods, especially multi-dimensional data visualization. Pixel-oriented techniques use each pixel of the display to visualize one data value and therefore allow the visualization of the largest amount of data possible. Pixel bar charts technique is to use the pixels within the bars to present the detailed information of the data records. We apply pixel bar charts in the analysis of real-world census data shows the wide applicability and usefulness of the technique. The thesis simplifies pixel bar charts and adapted it.The work done in the thesis can be simplified as following:1. Review the current visualization techniques, classify them into severalcategories, and select pixel-oriented visualization technique as my research focus.2. According the main factors of designing oriented-pixel visualization technology,the newly presented Pixel Bar Chart technology did a good job on visualizing large quantity data sets; thesis studied the algorithm and analyzes it.3. Implement Pixel Bar Chart technology and simplify it.4. Implement a Query-Independent visualization technique and study on the problem of dispersal of the pixels belonging to one data object in several sub windows.5. Implement an adapted pixel-oriented technique and compare with original technique.6. Implement sun icon visualization technique presented in the thesis.Through the research and practice, we find that we should choose different visualization techniques according to the characteristic of data, such as dimension, quantity and type of data sets, when visualizing data.