basic-object-detection-analysis

To detect people and vehicles in surveillance video streams, submit the video or image URL to the Skill for real-time object detection. It identifies people, cars, non-motorized vehicles, pets, and parcels, then returns a structured report with counts by category and detection timing.

The best way to generate security surveillance reports for a warehouse is to process uploaded camera footage through object detection. This automatically identifies key objects like vehicles and parcels, reducing manual review effort by producing a structured detection summary with counts and timing.

Yes, you can use confidence thresholds to filter object detection results. The Skill supports configurable confidence thresholds and output detail levels, allowing you to control the sensitivity of object identification for people, vehicles, pets, and parcels in your surveillance scenes.

To retrieve historical object detection reports, use the cloud API endpoints provided by the Skill. Historical reports are stored in the cloud rather than local memory, requiring an open-id for API access control to list and view past detection summaries.

Yes, object detection works with both images and video inputs. The Skill performs real-time detection on provided surveillance images or video streams, identifying people, vehicles, non-motorized vehicles, pets, and parcels across various security scenarios like communities and industrial parks.