mustafa
unread,Sep 18, 2012, 9:56:54 AM9/18/12Sign in to reply to author
Sign in to forward
You do not have permission to delete messages in this group
Either email addresses are anonymous for this group or you need the view member email addresses permission to view the original message
to android...@googlegroups.com
hi i want to implement an android app to handle face recogniton using PCA and my steps :
1- i have 3 images (100 * 100) built from them the matrix (10000 * 3) each image in column
Mat first = (Mat) this.trainImages.get(0);
this.total = first.cols() * first.rows();
// this matrix will have one column for each image
this.matrix = new Mat(this.total, this.trainImages.size(),
CvType.CV_32FC1);
for (int i = 0; i < this.trainImages.size(); i++) {
Mat matrixCol = this.matrix.col(i);
Mat trainImage = (Mat) this.trainImages.get(i);
trainImage.reshape(1, this.total).col(0).convertTo(matrixCol,
CvType.CV_32FC1);
}
2- calling Core.PCACompute(this.matrix, this.mean, this.eigenvectors,12);
3- now (here i have exception ) i want to project a new image onto the pca subspace by calling Core.PCAProject(m, this.mean, this.eigenvectors, projected);
where m is (10000 * 1) matrix it always give me an exception but when i projected my training 3 images matrix by this Core.PCAProject(this.matrix, this.mean, this.eigenvectors, this.projectedMatrix);
it didn't give an exception and return the projectedMatrix (10000 * 3) and where the calculated mean is (1*3) and the eigenvectors (3*3)
so where is the problem how i project new images to compute the similarity
thank