error in my program

sir, i want to classify testdata and traindata using knnclassify, I have 8 images per person total 80 images of dorsal hand vein, i made 10 classes and i used 3 images per person for testdata and 5 images for traindata, i extract minutie from the vein images. so, my problem is i got matrix of size 600 x 6 testdata and 1000 x 6 as traindata features am having the trainlabel size 50 X 1 when i specify this in knnclassify i got the error. predictlabel = knnclassify(reduced_testdata, reduced_traindata, train_label,3,'euclidean','nearest'); cp = classperf(test_label,predictlabel); cp.CorrectRate*100

[Walter Roberson]

Why is it x 6?

You should have something that is number of features in one direction, and number of images per person times number of people in the other direction. So that should be something by 30 for test, and by 50 for training.

[Balaji M. Sontakke]

Thanks sir