Table 4.

Summary of proposed framework.

Step 1. Import images $\left(X_{init}^{M\times H\times W}, y\right)$, where $M$ is the number of images, $H$ and $W$ are the number of pixels corresponding to the Height and Width of every image. Step 2. Compute Co-Occurrence, Run-length, Statistical and Contour features and extract new dataset $\left(X_{new}^{M\times N}, y\right)$, where $N$ is the number of new extracted features. Step 3. Train White Box Linear model LR with $\left(X^{M\times N}, y\right)$. Step 4. Define a weight threshold $d_{th}$ and compute most important features $K$. Step 7. For every new instance $X_{new}^{1\times N}$ and every feature $k \in K$ compute its critical values $xcri{t}_k$. Step 8. Verify explanation Conditions A, B and C. Step 9. Select and define the language of the explanation with respect to the targeted audience. Step 10. Create the Final Presentation explanation output.