Time of flight camera for welding machine vision
A machine-vision-assisted welding system comprises welding equipment, a time of Flight (ToF) camera operable to generate a three-dimensional depth map of a welding scene, digital image processing circuitry operable to extract welding information from the 3D depth map, and circuitry operable to control a function of the welding equipment based on the extracted welding information. The welding equipment may comprise, for example, arc welding equipment that forms an arc during a welding operation, and a light source of the ToF camera may emit light whose spectrum comprises a peak that is centered at a first wavelength, wherein the first wavelength is selected such that a power of the peak is at least a threshold amount above a power of light from the arc at the first wavelength.
1. A machine-vision-assisted welding system comprising:
welding equipment;
a first Time of Flight (ToF) camera operable to generate a three-dimensional (3D) depth map of a welding scene, wherein the first ToF camera is configured to capture a 3D map of an actual bead stack up;
digital image processing circuitry operable to extract welding information from said 3D depth map and compare the 3D map of the actual bead stack-up to a desired weld geometry; and
circuitry operable to control a function of said welding equipment based on said extracted welding information and said comparison.
2. The machine-vision-assisted welding system of claim 1 , wherein said welding equipment comprises an automation device.
3. The machine-vision-assisted welding system of claim 1 , wherein said welding equipment is operated by a human operator.
4. The machine-vision-assisted welding system of claim 1 , comprising:
an image sensor comprising a plurality of pixels, each of which is operable to convert light energy incident on it to an electrical signal; and
circuitry operable to, for each one of said pixels, measure time required for light to travel from a light source of said ToF camera to said welding scene and back to said one of said pixels.
5. The machine-vision-assisted welding system of claim 1 , wherein:
said welding equipment comprises arc welding equipment that forms an arc during a welding operation; and
a light source of said first ToF camera is configured to emit light whose spectrum comprises a peak that is centered at a first wavelength, wherein said first wavelength is selected such that a power of said peak, as measured at a particular pixel of an imager of said first ToF camera after reflection off of said welding scene, is at least a threshold amount above a power of light from said arc at said first wavelength, as measured at said particular pixel of said imager of said first ToF camera.
6. The machine-vision-assisted welding system of claim 1 , wherein:
said extracted welding information characterizes activity of one or both of a human and a machine in said welding scene; and
said function comprises activity tracking, cycle time improvement, and workflow streamlining.
7. The machine-vision-assisted welding system of claim 6 , wherein said activity comprises one or more of: welding operations, pre-welding operations and post-welding operations.
8. The machine-vision-assisted welding system of claim 1 , wherein:
said extracted welding information comprises activity of one or both of a human and a machine in said welding scene; and
said function comprises real-time assessment of risk of personal injury.
9. The machine-vision-assisted welding system of claim 1 , wherein said digital image processing circuitry uses data from sources other than said first ToF camera to interpret said 3D depth map.
10. The machine-vision-assisted welding system of claim 1 , comprising a second ToF camera, wherein said first ToF camera and said second ToF camera differ in one or more of: field of focus, mounting location, wavelength of light used, and targets viewed.
11. The machine-vision-assisted welding system of claim 1 , comprising a non-ToF camera configured to capture visible-spectrum images, wherein said information from said depth map and visible-spectrum images are combined to provide an augmented reality user interface.
12. The machine-vision-assisted welding system of claim 1 , comprising a non-ToF camera configured to capture infrared images, and being configured to use said infrared images to provide a temperature overlay on said 3D depth map.
13. The machine-vision-assisted welding system of claim 1 , wherein:
said extracted welding information comprises identification of one or more of: welding equipment, welding consumables, and welding operator; and
said function comprises one or more of: verification of compliance with a welding procedure specification (WPS), welder performance qualification (WPQ), procedure qualification record (PQR), and generation of alerts for error prevention.
14. The machine-vision-assisted welding system of claim 1 , wherein:
said extracted welding information identifies a sequence of welds being made on a workpiece in said welding scene; and
said function comprises one or both of:
indication, via a user interface of said welding equipment, of a location of a next weld in said sequence; and
one or both of generation of an alert and triggering of a lock out for prevention of wrong weld sequence.
15. The machine-vision-assisted welding system of claim 1 , wherein:
said extracted welding information comprises frequency and/or amplitude of surface oscillations of a weld puddle; and
said function comprises controlling said welding equipment based on said frequency and/or amplitude of said surface oscillations.
16. The machine-vision-assisted welding system of claim 1 , wherein:
said extracted welding information comprises joint fit-up condition of a workpiece in said welding scene; and
said function comprises adaptation of welding parameters used by said welding equipment based on said fit-up condition.
17. The machine-vision-assisted welding system of claim 1 , wherein:
said extracted welding information comprises actual joint seam location; and
said function comprises guiding a robot of said welding equipment to ensure a welding tool center point follows said joint seam.
18. The machine-vision-assisted welding system of claim 1 , wherein:
said extracted welding information indicates maintenance condition of said welding equipment; and
said function comprises one or both of:
preventative maintenance service; and
condition-based maintenance service tracking and alert generation.
19. The machine-vision-assisted welding system of claim 1 , wherein:
said extracted welding information indicates remaining stock of welding consumables; and
said function comprises generation of an alert to replenish a supply of said welding consumables.
20. The machine-vision-assisted welding system of claim 1 , wherein:
said extracted welding information indicates an identity and position of a welding workpiece in said welding scene; and
said function comprises automatic selection of a weld program to be used for welding said workpiece, based on said identity and position of said welding workpiece.
21. The machine-vision-assisted welding system of claim 1 , wherein:
said extracted welding information indicates a position and orientation of a welding torch in said welding scene; and
said function comprises automatic generation of a weld program to be used by an automated robot for welding said workpiece.
22. The machine-vision-assisted welding system of claim 1 , wherein:
said extracted welding information indicates an identity and position of a welding workpiece in said welding scene; and
said function comprises automatic selection of a weld program to be used for welding said workpiece, based on said identity and position of said welding workpiece.
23. The machine-vision-assisted welding system of claim 1 , wherein:
said extracted welding information indicates the dimensions of previous welds in a multi-pass weld or clad of the welding workpiece in said welding scene; and
said function comprises automatic planning of the remaining one or more passes to be used for multi-pass welding of the said workpiece, based on a previous one or more welding passes on said welding workpiece.
24. The machine-vision-assisted welding system of claim 1 , wherein:
said welding information indicates hand gestures made by a worker in said welding scene; and
said function comprises controlling said welding equipment based on said hand gestures.
25. The machine-vision-assisted welding system of claim 1 , wherein said function comprises acquisition of 3D depth data of said welding scene for machine learning and predictive analytics.
26. The machine-vision-assisted welding system of claim 1 , wherein:
said welding information comprises one or more of: weld presence, weld size, weld length, weld location, workpiece distortion, and weld appearance, and
said function comprises post-welding visual inspection of weld quality.