Museums and historical societies are increasingly adopting AI tools to enhance their collections, but they face unique challenges regarding accuracy and authenticity. I volunteer at a small local history museum that has thousands of glass plate negatives from the early 1900s, many of which have never been printed or properly cataloged. Scanning these negatives produces large format images, but they remain in black and white, which limits their appeal to casual visitors. We decided to experiment with colorizing a selection of street scenes from our town’s main avenue in 1910, and the Photo Colorizer produced results that were both beautiful and historically informative. The colors revealed details that had never been noticed before, such as the specific paint colors on storefronts and the varied hues of horses pulling carriages. Visitors now spend much more time looking at these colorized images than at the original black and white prints, and they ask better questions about daily life in the past. This increased engagement has justified the cost of digitizing the entire collection, and we plan to colorize hundreds more images for a special exhibition next year.
