Chapter 09 · Sustainability & AISustainability & AI

Computer vision

Meaning statusEstablishedSource recordDirect document linkedWhy these are different

Definition

The field of AI concerned with interpreting and extracting information from images and video — classification, detection, segmentation and tracking.

References

Science 2013Hansen et al. — High-resolution global maps of 21st-century forest cover change

This source provides part of the technical or institutional basis for the definition.

ICML 2021 / arXivRadford et al. — Learning Transferable Visual Models From Natural Language Supervision

This source supports the explanation of how the term is applied, measured or governed in practice.

Overview

What it means

Modern computer vision is dominated by deep learning applied to photographs, satellite scenes, camera traps and industrial cameras. Models convert pixels into structured observations: a species, a land-cover class, a methane plume, a defect.

How it is used

Environmental applications are extensive: forest-cover change detection, wildlife monitoring, crop assessment, waste sorting and infrastructure inspection. Global Forest Watch's deforestation alerts, built on Hansen et al.' s global forest maps, are a flagship example of vision at planetary scale.

Why it matters

Much of the natural world cannot be measured by sensors in the ground but can be photographed. Computer vision turns that imagery into evidence at scales no field team could match — making it foundational to environmental monitoring and MRV.