Incremental PCA-LDA Algorithm

In this paper a recursive algorithm of calculating the discriminant features of thePCA-LDA procedure is introduced. This algorithm computes the principalcomponents of a sequence of vectors incrementally without estimating thecovariance matrix (so covariance-free) and at the same time computing the lineardiscriminant directions along which the classes are well separated. Two majortechniques are used sequentially in a real time fashion in order to obtain the mostefficient and linearly discriminative components. This procedure is done bymerging the runs of two algorithms based on principal component analysis (PCA)and linear discriminant analysis (LDA) running sequentially. This algorithm isapplied to face recognition problem. Simulation results on different databasesshowed high average success rate of this algorithm compared to PCA and LDAalgorithms. The advantage of the incremental property of this algorithmcompared to the batch PCA-LDA is also shown.