Multi-sensor target recognition attribute reduction method and apparatus

Abstract

The present application provides a multi-sensor target recognition attribute reduction method and apparatus. The method includes the following steps that: sensor data are preprocessed, so that a plurality of target attribute data are obtained; for each target attribute datum, a K-means clustering method is used to determine a target clustering number corresponding to the target attribute data according to a preset rule on the basis of a preset category number parameter of target recognition, and data fuzzy processing is performed on a clustering result corresponding to the target clustering number; and a rough set algorithm is used to perform attribute reduction on a fuzzy processing result, so that an attribute reduction result is obtained. According to the multi-sensor target recognition attribute reduction method and apparatus of the present invention, a characteristic that the same attribute data of the same type of targets of the sensor data are just slightly different from each other, and the same attribute data of different types of targets of the sensor data are largely different from each other is considered; the K-means clustering method is adopted to perform clustering; the data fuzzy processing is performed on the clustering result; continuous data are discretized; and therefore, the limitations of a rough set in attribute reduction can be eliminated, and the recognition rate of fuzzy rough set-based target recognition can be improved.

Description

technical field [0001] The invention relates to the technical field of target recognition, in particular to a multi-sensor target recognition attribute reduction method and device. Background technique [0002] Rough set is a mathematical tool for studying imprecise and uncertain knowledge. The theory has been widely used in fields such as data mining, machine learning, process control, decision analysis and pattern recognition, and has achieved good results. Attribute reduction is an important topic in rough set theory. It is to derive the decision-making or classification rules of the problem through the simplification of knowledge under the premise of keeping the classification ability unchanged. Its significance is that redundant information can be deleted. [0003] The application of multi-sensors in the field of object recognition is a hot issue in knowledge discovery at this stage. Sensors can provide diverse, multi-dimensional, and real-time large amounts of data a...

AI Technical Summary

Problems solved by technology

However, the variety and multi-dimensional characteristics of sensor data and the nonlinearity of data make the traditional attribute reduction algorithm unable to work well.

The rough set theory can deal with this kind of irregular data well, but the data characteristics of the sensor data and the rough set's own characteristics limit the application range of the rough set on the sensor data

Images