EO/IR Sensor Integration for Target Identification

The wavelet transformation based multisensor integration approach represents a new method of combining advanced signal processing techniques to perform target identification. To utilize all available information, all operations involved in the sensor integration such as contour extraction, image registration, data fusion, clutter elimination, and target recognition are carried out at multiple resolution levels based on wavelet transforms. This leads to an increased ability to discrimanate targets and an increase in robustness due to redundance of information obtained. Unlike existing wavelet based approaches, this new approach computes the wavelet transform for each sensor image only once. Therefore, it provides not only computational efficiency but also a great degreee of robustness. It will be highly effective for identifying targets in rich clutter environment.