**Section 4**

**Application to Image Processing and Cognition Problems** 

240 Fuzzy Inference System – Theory and Applications

Zou, S. Huang, Y., wang,Y. Wang, J. & Zou, C. (2008). SVM learning from imbalanced data

*Comput., Sci* Hunan, China, 2008, pp.982-987.

by GA sampling for protein domain prediction. *Proc. of 9th Int. Congf. Joung* 

**11** 

*Italy* 

**Fuzzy Inference Systems Applied to** 

*Scuola Superiore S. Anna, TeCIP Institute, Pisa* 

**Image Classification in the Industrial Field** 

Silvia Cateni, Valentina Colla, Marco Vannucci and Alice Borselli

In the last years many industries have increased the exploitation of vision systems applied to several fields. This fact is basically due to the technological progress that has been reached by these systems: the reliability of vision systems allows the industries to achieve

Among the most important applications of vision systems in the industrial field there are robot positioning and driving, code reading, non contact measuring as well as quality

In particular quality controls are nowadays performed through vision systems in many industries; these systems guarantee reliability comparable and sometimes greater with respect to human operators, especially for a large quantity of products. In fact, vision-based automatic inspection systems allow to process a huge amount of data in a very small time

The vision systems are usually composed by several components: a set of cameras to capture the images, an illumination system and a system for the image processing and classification. These systems are able to capture images of a wide range of products in order to find defects which do not fit the quality standards of the considered industrial production process.

The present chapter deals with the application of Fuzzy Inference Systems (FIS) for the classification of images, once they have been pre-processed through suitable algorithms. One of the main reasons for exploiting a FIS to this aim lies in the possibility of approaching

 To translate the human experience (which is very often the reference starting point in industrial applications) through inference rules. The possibility of providing a methodological description in linguistic terms is a great advantage in problems that cannot be solved in terms of precise numerical relations, especially when only empirical

 To perform an eventual further adaptation of the rules of the inference machine by exploiting the available data, as a neural training algorithm can be applied to tune some

considerable cost savings in terms of both material and human resources.

the problem in a way similar to human reasoning. In fact a FIS allows:

To describe the problem in linguistic terms;

a priori knowledge is available.

parameters of the fuzzy classifier.

**1. Introduction** 

control and monitoring.

with respect to human performance.
