Pattern Recognition and Robotics

Pattern Recogniton In Robots

The problem of pattern recognition involves decision making in most of the robotics applications. The process of preprocessing in pattern recognition of signal/image processing is required to remove noise and redundant data. In the decision making approach, the process consists of data segmentation, feature extraction and feature selection/reduction as shown in Figure 1.

Figure 1.

Illustration of pattern recognition based robot system

After preprocessing of data, it is necessary to extract the features that will potentially help in decision making. Features contain information and the extraction of relevant features for decision making application depends upon the patterns and the number of decisions under consideration. Several approaches are available for extraction of features, selection of features and classification in different robotic applications.