(43a) Analyzing the Process Image Based on Visual Saliency and Visual Clutter

As an important type of process data,
images have been widely used in user-interface design and process monitoring because
they can perform non-invasive analysis on the processes or products with low
costs. However, an important issue with images is that they usually contain
redundant information which can interfere with searching for the target item.
For example, the process diagrams exhibited in the user-interface usually contain
a large amount of equipment such as tanks and columns, and it is difficult for
the user to find target items such as the reactor. A related topic is how to
measure the negative influence of excessive items on the target searching, and
based on the result we can improve the user-interface design by reducing the
number and changing dyes of items. Another example is related to the flare
analysis where the multivariate image analysis (MIA) has been established and
widely used [1]. Isolating the flare flames in the images from the environment
can be performed manually by visual detection. However, it is not quite
efficient. Because the flare flames usually have distinct perceptual quality
which makes them stand out from the environment, we can certainly apply
advanced methods from computer science community to perform such a task
automatically.

In this work, we apply methods including
the visual saliency detection [2] and clutter analysis [3] from computer science
community to analyze the process images to improve the user-interface design
and facilitate the flare flames detection.

Reference

[1] D.  ADDIN EN.REFLIST Castineira,  B.C. Rawlings, and T. F. Edgar.
Multivariate Image Analysis (MIA) for Industrial Flare Combustion Control.
Industrial& Engineering Chemistry Research, 51(39): 12642-12652, 2012.

[2] J. Harel, C. Koch, and P. Perona. Graph-based visual
saliency. In Advances
in neural information processing systems, 545-552. 2006.

[3] R. Rosenholtz,
Y. Li, L. Nakano. Measuring visual clutter. Journal of Vision,7(2):17, 2007.