Back to results
US 2026/0245215 A1Published application
Portable railroad grade crossing monitoring system
AssigneeUniversity of South Carolina
InventorYu Qian
Life of this patent
Estimating the term…
Abstract
Described herein is a portable edge-computing system, and methods for use and making same, to detect moving and stationed objects at railroad-highway grade crossing areas in order to detect foreign objects that do not belong at or near the specific location to identify potential threats, especially for detecting pedestrians and vehicles at railroad-highway grade crossing areas.
Claims20
A portable railroad grade crossing monitoring system comprising:
- at least one input head configured to both identify and discern foreground objects within at least one visual scene;
- at least three processors configured to act independently from one another comprising;
- a first processor configured to process foreground detection;
- a second processor configured to process object classification;
- a third processor configured to process motion tracking;
- and
- wherein each processor is assigned to a separate central processing unit core;
- the at least one input head further configured to compare at least one current frame and at least one static background and further configured to isolate at least one dynamic object or at least one static object in a foreground of the at least one visual scene to detect at least one moving object against at least one background image.
5 more elements
A method for monitoring a railroad grade crossing comprising:
- positioning at least one device comprising at least one input head at a railroad grade crossing;
- configuring the at least one input head to harmonize weight between at least one video frame input and at least one background image input;
- wherein the at least one input head transforms both the at least one video frame input and at least one background image input into at least one first feature map and at least one second feature map wherein the at least one first feature map and the at least one second feature map are identical in size;
- and
- configuring the at least one device to comprise at least three processors to act independently from one another including;
- a first processor configured to process foreground detection;
- a second processor configured to process object classification;
- a third processor configured to process motion tracking;
- and
- wherein each processor is assigned to a separate central processing unit core;
7 more elements
Dependencies are inferred from the claim text.