Performance measures for pattern recognition in a wavelet joint transform correlator

摘要

The present paper shows the results of the implementation of a correlation Digital processed Joint Wavelets for gray level images target recognition applications. The aim is to study and characterize the different families of two-dimensional wavelets that are available and build a system of recognition of images it becomes operational, flexible and reliable. The resulting methodology of this study becomes a useful tool to implement in the future, in which the evolution of technology will provide greater reliability and algorithms will allow the design of correlation more compact, fast and versatile with multiple applications. The correlation was tested in image recognition, which also involved images that suffered distortion type Gaussian noise in order to verify its robustness to the ability of discrimination correlation Transformed Joint (JTC) classics.