amine's profile - activity

I've been working on an AAM model using opencv for quite a while now and have been able to extract 2D coordinates from an image , and using POSIT (solvePnP) I have been able to extract the translation and rotation vector from the image , my issue is , that now I do not know how to reverse this translation and rotation to got a centralized normalized point set . Thanks in advance.

I have an array of point sets to which I have assigned a value but after training everytime I try to predict a new dataset I recieve the same value that seems to be the arithmetical average of my training responses here's a code sample for my training:

machine = cv2.svm()
trainData = numpy.array([[x1,y1,x2,y2...xn,yn],[x1,y1,x2,y2...xn,yn],...,[x1,y1,x2,y2...xn,yn]],dtype= numpy.float32)
results =  numpy.array([res1,res2,res3,...,resn],dtype = numpy.float32)
machine.train(trainData,results,dict( kernel_type = cv2.SVM_LINEAR, svm_type = cv2.SVM_C_SVC,C = 1 ))

I would be thankful if anybody pointed to me my mistakes ,