Question

The overall image classification accuracy (in percentage) calculated from the following error matrix is __________________ (in integer).

Hint:

The overall image classification accuracy is calculated by dividing the sum of diagonal elements by the total sum of the matrix and multiplying by 100.

Answer 1

Correct Answer: 82

To calculate the overall image classification accuracy, we use the following formula: \[ \text{Overall Accuracy} = \frac{\text{Sum of the diagonal elements}}{\text{Sum of all elements}} \times 100. \] From the error matrix: \[ \begin{array}{|c|c|c|c|c|} \hline \text{Thematic Map Classes} & \text{SOIL} & \text{WATER} & \text{CROP} & \text{Total}
\hline \text{Ground Truth Classes} & \text{SOIL} & 40 & 1 & 4 & 45
\text{WATER} & 7 & 25 & 3 & 35
\text{CROP} & 1 & 2 & 17 & 20
\hline \text{Total} & 48 & 28 & 24 & 100
\hline \end{array} \] The diagonal elements are 40 (SOIL), 25 (WATER), and 17 (CROP). The sum of these diagonal elements is: \[ \text{Sum of diagonal elements} = 40 + 25 + 17 = 82. \] The total number of elements in the matrix is: \[ \text{Sum of all elements} = 48 + 28 + 24 = 100. \] Thus, the overall accuracy is: \[ \text{Overall Accuracy} = \frac{82}{100} \times 100 = 82%. \] Hence, the overall classification accuracy is 82.