Multiple Choice
Which of the following is not a potential problem with unsupervised classification?
A) some spectral classes may be meaningless
B) a training area may contain a mix of different thematic classes
C) reference data used for assigning classes by the analyst may contain errors
D) a single thematic class may be split among two spectral classes
Correct Answer:
Verified
Correct Answer:
Verified
Q1: The number of pixels used in training
Q2: Georeferencing an image to geographical coordinates (latitude
Q3: Image classification is based on<br>A) tone or
Q4: With the classification error matrix below,
Q5: Cubic convolution is a geometric correction method,
Q7: On a colour infrared image (or photo),<br>A)
Q8: Microwaves are the longest wavelengths used for
Q9: With linear stretching,<br>A) the image would appear
Q10: In a near IR image, forest appears<br>A)
Q11: In supervised classification, training areas are<br>A) representative