Page 4 of the DSE Video Surveillance knowledge center, with source-backed guidance and practical next steps.
Practical guidance for camera, recorder, video-management, evidence, firmware, and surveillance-network decisions.
Wind, traffic, doors, machinery, and flexible mounts can turn a high-resolution or long-zoom camera into a moving image source. Correct the mechanics, select stabilization deliberately, and prove the recorded result under load.
Read the checklist
Thermal cameras can reveal people and vehicles through darkness and difficult backgrounds, but thermal imagery generally cannot identify a person. Design detection, visual verification, analytics, and response as separate tested functions.
Read the guide
A dome camera can look clean by day and become foggy, mirrored, or washed out when its infrared LEDs switch on. Night acceptance must test nearby surfaces, dome condition, lens seals, weather, and external light.
Read the checklist
Rolling and global shutters expose a scene differently. Select between them only after defining target speed, direction, exposure time, illumination, and the image detail the operator or analytic must preserve.
Read the guide
Detection, observation, recognition, and identification require different image detail. Define the task and target distance first, use pixel density as a planning aid, then prove the recorded result in the real scene.
Read the guide
WDR specifications do not prove a usable image at a bright entrance, garage, tunnel, or reflective lobby. Test the installed camera through changing light and motion, then preserve the configuration and evidence.
Read the checklist
License-plate recognition starts with a readable captured plate. Prove the camera at the real lane angle, vehicle speed, capture distance, day/night lighting, and recording settings before trusting the recognition result or gate action.
Read the playbook
A stream can be clear and still arrive too late. Measure sensor-to-screen delay through the production camera, network, VMS, decoder, workstation, and display—under normal and stressed conditions—before live video supports a time-sensitive response.
Read the checklist
Radar can add reliable movement, position, and speed data where light, fog, shadows, or privacy limit video. Its blind spots, reflections, classification limits, zones, and response integrations still need real-scene acceptance testing.
Read the checklist
Vehicle risk is shaped by approach speed, geometry, pedestrian exposure, deliveries, traffic operations, emergency access, and site use. Use CISA’s Plan-Prevent-Protect framework to define the problem before selecting a barrier.
Read the guide