Palmaesthetica is a generative art project and collection that uses algorithms to create beautiful images. It explores the ability of algorithms to endlessly generate permutations in composition and colour and how this affects the viewer.
Using nature and the palm trees of the Pacific as inspiration, it aims to meet the challenge of using algorithms to generate recognisable scenes, but that are at the same time not quite what we expect.
The graphic style focuses attention upon the juxtaposition of forms and space and the moments of interest, calm and happiness they can create.
Above all, I hope that every image will be enjoyed.
About the Project
The project or process that became Palmaesthetica began in August 2023. I was fascinated by the idea that the essence of a style of art, or a particular recognisable scene or composition, could be captured in an algorithm, and then endlessly permutated. In effect, you would be able to generate in moments what could traditionally have been the work of a lifetime.
To progress this idea, I set myself the challenge of abstracting a recognisable subject into a set of algorithms that could be used to generate a collection. It had to be visually appealing, unpredictable and allow for a large variety of interesting pieces to be generated. However, at the same time, it would not be entirely random and had to be directed by an intention. In short, it had to be a meaningful test of the ideas and not a trivial creation.
After experimenting for a few months, I settled on the well known subject of a palm tree because it leant itself to both naturalistic and also more abstract or graphic design based presentations. I found the more I distilled the scene into it’s essence, the more the quality of the algorithms became important, and so I could improve them. This further revealed that small changes, such as colour, had profound affects upon the look and feel of the resulting work. After 8 months of evolution I now have something I am happy to share with you.
The Technology
Every picture and element within the picture is generated by handcrafted algorithms that use the p5.js open source library to create the images. The principal challenge has been to automatically and consistently generate images with visual appeal. The initial algorithms capture the high level artistic intent, and use randomisation to enable the creation of a near limitless number of truly unique images. Automated curation algorithms then discard images that do not meet the original vision.
Each picture comes with a comprehensive set of metadata that describes it’s visual elements. However, the most technically significant value is the unique “Fingerprint” that identifies the image and from which every aspect can be derived.
The Details
Each unique piece in the collection has the following metadata characteristics. These are randomly decided for each image. While these can be used to assess the relative rarity of an image, each is visually unique and the appearance including placement of trees and their shapes must be assessed by sight.
1) Aspect. One of: UHD Portrait, UHD Landscape, Square (2160×2160), WQHD+ Portrait, WQHD+ Landscape
2) Palette. One of: Natural, Neon, Gold White, Gold Black, Sunset
3) Outline. Whether the piece includes outlined picture elements
4) Composer. Each image is composed using one of 7 different techniques
5) Number of Palms. Between 1 and 4. Limited in order to create interesting juxtapositions between trees
6) Rainbow Palms. A very small number of pieces include a rainbow coloured palm
7) Left Facing and Right Facing numbers of palms
8) Concave and Convex numbers of palms
9) Fronds. The total number of palm fronds in the image
10) Leaflets. The total number of leaves across all palms in the image
GN Miller Art 