Definition and goals of computer vision

Summary

Computer vision is an interdisciplinary field focused on how computers can gain high-level understanding from digital images or video. From an engineering perspective, it aims to automate tasks that the human visual system can do; from a scientific perspective, it studies the theory behind artificial systems that extract information from images. In practice, computer vision involves automatically extracting, analyzing, and understanding useful information from one image or a sequence of images, producing numerical or symbolic outputs such as decisions. A central goal of computer vision is “understanding,” meaning transforming visual data into descriptions of the world that are meaningful to reasoning processes and can support appropriate action. This often requires disentangling symbolic information from raw image data using models informed by geometry, physics, statistics, and learning theory. The field also distinguishes between its scientific role (developing theory) and its technological role (building computer vision systems), with machine vision increasingly overlapping in modern usage. Computer vision systems are designed to handle many forms of visual input (e.g., single images, video sequences, multi-camera views, 3D point clouds, and medical scans) and to support a wide range of tasks such as recognition, motion analysis, scene reconstruction, and image restoration. Overall, the discipline provides both the conceptual foundations and the practical methods for turning visual observations into actionable information.

Key Points

• 1

  Computer vision automates the extraction, analysis, and understanding of useful information from images or video.

• 2

  “Understanding” means converting visual data into world descriptions that support reasoning and action.

• 3

  The field combines scientific theory (how information can be extracted) with technological practice (building vision systems) and spans many input types and tasks.