A Computer Vision Framework for Estimating Surface Habitability from Mars Using Convolutional Analysis

Identifying signs of life in extraterrestrial environments is one of the growing challenges in planetary science. Conventional approaches to detecting habitability rely heavily on direct contact with biosignatures or geological analyses, but limited data and mission costs hold back such methods. This work introduces a computer vision-based pipeline that analyzes planetary surface images to determine a livability index that estimates the photographed terrain’s likelihood to support life. The data is from the Mars Surface Image Dataset (collected by NASA’s Curiosity Rover) and the Mars Handlens Analog Database, archived at the Planetary Data System (PDS) Geosciences Node of Washington University in St. Louis. Unlike existing binary classifiers that only distinguish between planetary sources, our method utilizes feature extraction to evaluate environments in terms of their habitability. By combining convolutional features with descriptors of texture, hue, and structure, we demonstrate that our model can move beyond simple classification and instead generate interpretable, probabilistic estimates of habitability. With a livability index, we can focus on key habitability features and use them to guide targeted exploration to better search for life beyond Earth.