MEDIAN IMAGERY
What I call "median images", are more accurately termed "average images". I create them using a simple algorithm that adds together many photographs and produces a resulting image that is, in affect, an average representation of the input photographs. When photographs contain some commonalities and are averaged, patterns may then emerge that reflect regularities that exist across all photographs. For instance, by averaging hundreds of images of a temperate deciduous forest, common patterns of vertically orientated structures (tree stems) and a vertical colour profile emerge. In a way, the resulting median image shows features that are shared by all members of the scene or environment (i.e., input photographs).
I focus on creating median images of natural landscapes and ecosystems. I like to think that each resulting median image is a portrait of that ecosystem and captures some holistic beauty and representation of its complexity. Typically, the median images I create are composed of 60 - 200 photographs of a landscape or ecosystem, although some are created from over 1000 photographs. I was initially inspired to create median images after exploring the innovative and impressive work done by the computer vision researcher, A. TorraIba and members of his lab at MIT.
I focus on creating median images of natural landscapes and ecosystems. I like to think that each resulting median image is a portrait of that ecosystem and captures some holistic beauty and representation of its complexity. Typically, the median images I create are composed of 60 - 200 photographs of a landscape or ecosystem, although some are created from over 1000 photographs. I was initially inspired to create median images after exploring the innovative and impressive work done by the computer vision researcher, A. TorraIba and members of his lab at MIT.
