[x] Add option to align images based on the mean image of the set of all images (original and copies). This ensures optimal comparison between images to enalbe clustering
[x] Preprocessing steps in terms of sharpness, contrast and brightness from the alignment page should also be available here, but now enabling to set them to the mean levels of sharpness, contrast and brightness
[x] Add option to convert all images to gray to ensure uniformity (i.e., so that color definately won't play a role in the cluster process)
[ ] Use the following set of measures as the basis for clustering:
Structural Similarity Metrics: SSIM, MS-SSIM, CW-SSIM, IW-SSIM
Perceptual Similarity Metrics: FSIM, VIF, LPIPS, PieAPP
Gradient-Based Metrics: GMSD, MDSI, DISTS
Saliency-Based Metrics: SR-SIM, VSI
[x] Add brushstroke technique as an additional component to cluster on
[ ] Add evaluation metrics on succes-rate of cluster process (how homogenous are the images inside a cluster? How distinct are the clusters?
[ ] Add option to provide background knowledge on which images belong to each other to assess the accuracy and correctness of the cluster process
[x] Plot clusters of images to visually inspect their correspondences (together with information on year of painting?)
[x] Add clustimage functionality in terms of scatterplot and dendrogram, and using its functionality to derive the optimal number of clusters
[x] Add scatterplot to visualize clustering results (independent from clustimage)
[x] Add dendrogram to visualize clustering results (independent from clustimage)
[x] Add heatmap to visualize matrix of structural similarity values (like feature points matrix). Present matrix after equalisation of parameters, alignment procedure and selection of cluster method and index, but before plotting the clusters and associated images
Preprocessing: input for clustering