1. Explain the following phrases used in geometric correction: (15 points) Geometric rectification and registration; GCP; RMSE; Re-sampling Methods; ISODATA. 2. Explain and compare three image classification decision rules: Minimum distance to mean, maximum likelihood, and parallelepiped (15 points) 3. Based on the error matrix (accuracy report) below, calculate the following (18 points): Error Matrix for Example Supervised Classification Reference Data Classified Data Corn Soybean Grass Tree Water Total Row Corn 156 8 10 15 1 190 Soybean 36 138 25 0 206 Grass 33 19 95 2 157 25 15 20 0 120 0 0 0 17 17 Tree Water Total Column 250 180 150 7 8 60 0 90 20 690 a. The producer's accuracy and the omission error for each of the classes b. The user's accuracy and the commission error for each of the classes c. Overall accuracy of this image classification d. Interpret and compare the accuracies of each class (eg. which is higher or lower as well as possible reasons) e. Write down the number of reference data for each class f. Describe the meanings of the produce's (or analyst's) accuracy and the user's accuracy.