Some pre-processing to extract the main color of each book cover in this dataset, before visualizing the data in this Observable notebook.
The extracted colors are eventually rounded to a set of eleven distinct colors, using CIEDE2000 Color Distance and then grouped and visualized, by genre. Might expand this a little in the future, if I find some more interesting patterns, but pretty simple and straightforward as it is.
Some future areas of exploration would be:
- Scraping the release years or each book in the dataset and looking for patterns between decade of release and use of colors.
- Extracting title size data or other data about text on the cover (e.g. how many non-title words) and looking for correlations with genre/release year.