Apparatus and method for inspecting assembly hole of vehicle
AssigneesHyundai Motor Company, Kia Corporation
InventorsSung Chul Yang, Min Ook Park
Life of this patent
Estimating the term…
Abstract
An apparatus and a method for inspecting an assembly hole of a vehicle are provided. The apparatus may comprise a sensor device configured to obtain an image of a part-fastening area and a processor configured to detect an assembly hole of a vehicle body in the image and configured to perform artificial intelligence (AI) learning of a shape of the assembly hole to determine whether the part-fastening area is defective.
Claims8
An apparatus for inspecting an assembly hole of a vehicle, the apparatus comprising:
- a sensor device configured to obtain an image of a part-fastening area;
- and
- a processor configured to:
- detect an assembly hole of a vehicle body in the image;
- and
- perform artificial intelligence (AI) learning of a shape of the assembly hole to determine whether the part-fastening area is defective;
- wherein the processor is further configured to:
- detect an overlap area where a first hole formed in a first panel of the part-fastening area and a second hole formed in a second panel of the part-fastening area overlap with each other;
- determine whether the first hole or the second hole corresponds to a circular hole used to mount a chassis;
- and
- determine that the part-fastening area is defective when the area where the first hole and the second hole overlap with each other is greater than or equal to a threshold size.
An apparatus for inspecting an assembly hole of a vehicle, the apparatus comprising:
- a sensor device configured to obtain an image of a part-fastening area;
- and
- a processor configured to:
- detect an assembly hole of a vehicle body in the image;
- and
- perform artificial intelligence (AI) learning of a shape of the assembly hole to determine whether the part-fastening area is defective;
- wherein the processor is further configured to;
- detect an overlap area where a first hole formed in a first panel of the part-fastening area and a second hole formed in a second panel of the part-fastening area overlap with each other;
- determine whether the first hole or the second hole is a square hole used to assemble a front glass;
- and
- determine that the part-fastening area is defective when a length of any segment connecting both sides facing each other in the square hole is less than a predetermined threshold length.
An apparatus for inspecting an assembly hole of a vehicle, the apparatus comprising:
- a sensor device configured to obtain an image of a part-fastening area;
- and
- a processor configured to:
- detect an assembly hole of a vehicle body in the image;
- and
- perform artificial intelligence (AI) learning of a shape of the assembly hole to determine whether the part-fastening area is defective;
- wherein the processor is further configured to:
- detect an overlap area where a first hole formed in a first panel of the part-fastening area and a second hole formed in a second panel of the part-fastening area overlap with each other;
- determine whether the first hole or the second hole is a notch hole used to mount a rear quarter glass, and
- determine that the part-fastening area is defective when a gap where the notch hole and another panel in which the notch hole is not formed overlap with each other is greater than or equal to a predetermined threshold.
A method for inspecting an assembly hole of a vehicle, the method comprising:
- obtaining an image of a part-fastening area;
- detecting an overlap area where a first hole formed in a first panel of the part-fastening area and a second hole formed in a second panel of the part-fastening area overlap with each other in the image;
- determining whether the first hole of the second hole is a circular hole;
- and
- performing artificial intelligence (AI) learning of a shape of the assembly hole;
- and
- determining that the part-fastening area is defective when the area where the first hole and the second hole overlap with each other is greater than or equal to a threshold size.
A method for inspecting an assembly hole of a vehicle, the method comprising:
- obtaining an image of a part-fastening area;
- detecting an overlap area where a first hole formed in a first panel of the part-fastening area and a second hole formed in a second panel of the part-fastening area overlap with each other in the image;
- determining whether the first hole or the second hole is a square hole used to assemble a front glass;
- performing artificial intelligence (AI) learning of a shape of the assembly hole;
- and
- determining that the part-fastening area is defective when a length of any segment connecting both sides facing each other in the square hole is less than a predetermined threshold length.
A method for inspecting an assembly hole of a vehicle, the method comprising:
- obtaining an image of a part-fastening area;
- detecting an overlap area where a first hole formed in a first panel of the part-fastening area and a second hole formed in a second panel of the part-fastening area overlap with each other in the image;
- determining whether the first hole or the second hole is a notch hole;
- and
- determining that the part-fastening area is defective when a gap where the notch hole and another panel in which the notch hole is not formed overlap with each other is greater than or equal to a predetermined threshold.
Dependencies are inferred from the claim text.