Rail feature identification system
AssigneeHarsco Technologies LLC
InventorJavier Fernandez
Life of this patent
Estimating the term…
Abstract
The present disclosure includes systems, devices, and methods for identifying, detecting, and/or tracking rail features. In some aspects, a system includes a camera and a computer having at least one memory, at least one processor configured to receive a plurality of images from the camera, and for each of the images: assigning a location identifier and identifying one or more rail features that correspond to one of a plurality of predetermined rail features. In some systems, the at least one processor is configured to determine a location of each of the one or more identified rail features.
Claims20
A rail vehicle comprising:
- a frame coupled to a plurality of rail wheels configured to move along a rail track;
- and
- a railroad track feature detection system coupled to the frame, the feature detection system comprising:
- a camera coupled to the frame and positioned such that the camera is configured to capture a plurality of images of a rail track as the rail vehicle moves along the rail track;
- and
- a computer having:
- at least one memory comprising instructions;
- and
- at least one processor configured to execute the instructions, wherein the instructions cause the at least one processor to:
- receive the plurality of images, and for each of the images:
- assign a location identifier;
- detect one or more rail features;
- identify the one or more rail features as a relevant rail feature from a dataset of predetermined rail features;
- and
- assign a feature identifier to each identified rail feature;
- and
- based on the location identifier and the feature identifier:
- determine a geographical location of the one or rail more features;
- and
- track the one or more rail features.
A railroad track feature detection system comprising:
- an imaging device configured to be coupled to a frame of a rail vehicle and positioned such that the imaging device is configured to capture a plurality of images of a rail track as the rail vehicle moves along the rail track;
- and
- a computer having:
- at least one memory comprising instructions, at least a portion of the instructions configured as a neural network;
- and
- at least one processor configured to execute the instructions, wherein the instructions cause the at least one processor to perform one or more operations in real-time;
- wherein the one or more operations include:
- receiving the plurality of images, and for each of the images:
- assigning a location identifier;
- and
- identifying one or more rail features that correspond to one of a plurality of predetermined rail features;
- and
- determining a location of each of the one or more identified rail features.
A method for identifying rail features using machine learning, the method comprising:
- obtaining, by one or more processors, first image data corresponding to database of a plurality of images of a rail track having one or more rail features;
- performing, by the one or more processors, object detection on at least a portion of the first image data to convert the at least a portion of the first image data to training data,
- training, by the one or more processors, one or more machine learning (ML) models based on the training data to configure the one or more ML models to identify predetermined rail features in subsequent image data;
- wherein the predetermined rail features includes a rail, a rail tie, a tie plate, an anchor, a spike, a spike hole defined by the tie plate, a joint bar, and/or a joint bar bolt.
Dependencies are inferred from the claim text.