What is Unusual Motion Detection?

Avigilon Unusual Motion Detection (UMD) is an advanced AI technology designed to enhance security by identifying and flagging atypical motion events within a camera’s field of view. Unlike traditional motion detection systems, UMD does not require predefined rules or setup. Instead, it continuously learns what typical activity looks like in a scene and then detects and flags motions that deviate from this norm.

Key Features of Avigilon Unusual Motion Detection

Continuous Learning:
UMD continuously analyzes motion-based activity to establish a model of typical behavior in a scene. Over time, it learns what constitutes normal motion and can then identify and flag unusual movements that deviate from this learned model.

Automation and Efficiency:
By automating the detection of unusual motion, UMD helps security operators filter through large amounts of recorded video more efficiently. This reduces the time required to review footage, allowing operators to focus on events that may need further investigation.

Focus of Attention Interface:
UMD leverages the Focus of Attention (FOA) interface in the Avigilon Control Center (ACC) software. This interface visually highlights atypical events, ensuring that critical incidents do not go unnoticed.

Embedded on H5 Platform Cameras:
UMD technology is embedded in Avigilon’s H5A and H5SL camera lines, making it accessible on both high-end and entry-level cameras. This integration provides powerful AI capabilities across a range of camera models.

Faster Search Capabilities:
The technology enables faster search through recorded video by flagging unusual events, which helps reduce hours of work to minutes. This feature is particularly useful for quickly identifying and responding to potential security threats.

Comparison with Unusual Activity Detection (UAD):
While UMD focuses on detecting anomalies in general motion within a scene, Unusual Activity Detection (UAD) is object-aware and detects atypical behaviors of specific classified objects, such as people and vehicles. UMD is suitable for environments where motion patterns are generally consistent and predictable, whereas UAD is more targeted towards monitoring the behavior of distinct objects.