Searching in CCTV: effects of organisation in the multiplex

Tatler, Benjamin W. · 2021 · Crossref

DOI: 10.1186/s41235-021-00277-2

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study investigates how the arrangement of camera feeds in a CCTV multiplex affects an operator’s ability to locate a specific scene, addressing the cognitive demands of surveillance. While arranging feeds by geographical proximity is a common recommendation, empirical evidence for its benefit is lacking. The authors propose an alternative organizing principle based on visual search theory: grouping scenes by semantic similarity (e.g., city centers, suburbs, traffic). The research aims to determine if semantic grouping facilitates search speed and whether this benefit depends on physical separation between scenes or the operator’s awareness of the arrangement. The study employed three experiments using a laboratory paradigm that simulated reactive surveillance. Participants searched for a target scene within a multiplex of 27 static images captured from real CCTV cameras in Dundee, Scotland. The images were categorized into three semantic groups. Experiment 1 used a 2x2 repeated-measures design manipulating the presence of gaps (borders) between scenes and the organization of scenes (mixed vs. semantic grouping), with participants unaware of the arrangement. Experiment 2 replicated this design but informed participants of the semantic grouping. Experiment 3 introduced varying target frequencies across categories to mimic the unequal distribution of crime in real-world settings. Participants identified a target scene and performed a discrimination task (identifying a T or L) to ensure active attention. The results demonstrated that searching for a target scene was significantly faster when scenes were arranged by semantic category compared to a mixed arrangement. In Experiment 1, where participants were unaware of the grouping, this benefit was observed only when scenes were separated by gaps. In Experiment 2, where participants were informed of the semantic structure, the benefit of semantic grouping persisted regardless of whether gaps were present. Experiment 3 revealed that when target frequency varied, the advantage of semantic organization was enhanced for the most frequently searched category, without a statistically significant cost for rarely searched categories. These findings suggest that organizing CCTV multiplexes by semantic category is an effective strategy to assist operators in locating specific scenes. The study extends the understanding of visual search by demonstrating that between-scene semantic relationships can guide attention, similar to within-scene semantics. The results imply that modern software-controlled data walls should prioritize semantic grouping over geographical proximity to reduce cognitive load. Furthermore, the data indicate that while physical separation between feeds aids search when the arrangement is implicit, explicit knowledge of the semantic structure allows operators to benefit from the grouping even without physical gaps. This provides actionable guidance for designing more efficient surveillance interfaces.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• visual search