how to test my patternnet for input image?

i m doing satellite image classification on the basis of NDVI values. image dimension is 100 by 100. i have 10 samples and i want to classify image into three categories. After extracting pixels of each category from image, i selected 30 pixels from each category for input.

so now my input - [20 30]

         target-[3 30]

IS my input and target matrix correct?

i have trained it using pattern net. after training now i want to give some image as input to test whether it is classified correctly or not.

so now what should be the dimensions of input image?

also how would i use the net work to display classified image?

plz help