System for Performing Computer-Assisted Image Analysis of Welds and Related Methods
A system for performing computer-assisted image analysis of welds and related methods is disclosed. Digital images are captured at a worksite and sent to a remote image analysis system of a weld analytics system that analyzes images to determine whether the images conform to weld specifications. The remote image analysis system may be trained by artificial intelligence or machine learning.
1 . An image analysis system, comprising:
a computer terminal and an imaging device connected to the computer terminal; and
a weld analytics system located remotely from the computer terminal and imaging device;
wherein the weld analytics system and computer terminal are configured to exchange data with each other over a communication network;
wherein the computer terminal comprises a processor configured to:
receive a digital image;
perform a first determination as to determine whether a weld shown in the digital image conforms to weld specifications;
generate a digital ticket in response to the first determination, the digital ticket comprising data comprising a unique image identifier (ID) and a first performance indicator representing a determination as to whether the weld shown in the digital image conforms to the weld specifications; and
send the digital image and the digital ticket to the weld analytics system;
wherein the weld analytics system comprises a repository and a server, the server comprising a processor configured to:
receive the digital image and the digital ticket;
store the digital image in the repository;
perform a second determination as to determine whether the weld shown in the digital image conforms to weld specifications; and
generate and store a weld record in the repository in response to the second determination, wherein the weld record comprises data comprising the unique image ID and the first performance indicator and/or a second performance indicator representing a determination as to whether the weld shown in the digital image conforms to the weld specifications.
2 . The image analysis system of claim 1 , wherein the processor of the computer terminal is further configured to:
determine whether the digital image conforms to image quality specifications.
3 . The image analysis system of claim 2 , wherein the processor of the computer terminal only performs the first determination as to determine whether the weld shown in the digital image conforms to weld specifications in response to a determination that a digital image conforms to the image quality specifications.
4 . The image analysis system of claim 2 , wherein the processor of the computer terminal is further configured to:
generate an alert in response to a determination that the digital image does not conform to image specifications, wherein the alert is displayed on a display of the computer terminal.
5 . The image analysis system of claim 2 , wherein the image quality specifications specify a plurality of image qualities comprising one or more of an image resolution, an exposure level, a focus level, a contrast level, and a size of the weld in the digital image.
6 . The image analysis system of claim 2 , wherein the processor of the weld analytics system is further configured to:
generate a digital ticket for each image that conforms to the image quality specifications, wherein the digital ticket comprises the first performance indicator and a unique ID to identify the respective image.
7 . The image analysis system of claim 6 , wherein the digital ticket further comprises one or more of a date the image was taken, a time the image was taken, a location at which the image was taken via telemetry, a welder who made the weld, a unique customer ID, and a unique job ID.
8 . The image analysis system of claim 7 , wherein the location is a geolocation which may include a description and/or a Global Navigation Satellite System (GNSS) coordinate.
9 . The image analysis system of claim 6 , wherein the digital ticket further comprises, in response to a determination that the weld shown in the digital image does not conform to weld specifications, one or more of a location of a flaw or defect in the weld, a type of a flaw or defect in the weld, a description of the flaw or defect, a description of the weld in machine, and a likelihood that the weld is conforming.
10 . The image analysis system of claim 6 , wherein the digital ticket is generated in a human readable format and encoded with eXtensible Markup Language (XML) extractable in downstream processing.
11 . The image analysis system of claim 1 , wherein the processor of the weld analytics system is further configured to:
generate a notification in response to a determination that the weld shown in the digital image does not conform to weld specifications.
12 . The image analysis system of claim 11 , wherein the notification is an electronic message sent to one or more designated recipients, wherein the notification preferably includes information about the weld extracted from a weld record, and wherein the notification preferably includes a copy of the digital image of the weld and/or the corresponding digital ticket.
13 . The image analysis system of claim 1 , wherein the processor of the computer terminal is further configured to:
determine a likelihood that the weld shown in the image conforms to the weld specifications.
14 . The image analysis system of claim 1 , wherein the digital image is a real-time radiography (RTR) image.
15 . The image analysis system of claim 1 , wherein the imaging device is an ultrasonic imaging device and the digital image is an ultrasonic image.
16 . The image analysis system of claim 1 , wherein the imaging device is a computed radiography device or digital radiography imaging device and the digital image is a radiographic image.
17 . The image analysis system of claim 1 , wherein the weld analytics system has been trained by artificial intelligence (AI) or machine learning (ML).
18 . The image analysis system of claim 17 , wherein the weld analytics system comprises an image classifier trained to classify images of welds by the type of weld and/or whether an image of a weld shows a weld conforming to weld specifications corresponding to the type of weld.
19 . The image analysis system of claim 18 , wherein the image classifier comprises at least one neural network.
20 . The image analysis system of claim 1 , wherein the processor of the weld analytics system is further configured to:
generate a new job record in the repository in response to a determination that the weld is non-conforming.
21 . The image analysis system of claim 20 , wherein the processor of the weld analytics system is further configured to:
generate a work order for repairing or fixing the non-conforming weld, wherein the work order preferably includes information generated from the corresponding weld record, wherein the work order preferably includes a technician and/or welder to perform the work, and a date on which the work is to be performed; and
electronically send the work order to one or more users.
22 . The image analysis system of claim 21 , wherein the work order comprises at least a scheduled date on which the work is to be performed, wherein the scheduled date is based on one or more of an importance of the weld, a customer and/or job, and construction due dates and/or timelines/schedule.
23 . The image analysis system of claim 22 , wherein the work order further comprises a technician and/or welder to perform the work, wherein the technician and/or welder is based on one or more of the welder who formed the non-conforming weld, a location of a flaw or defect in the weld, a type of flaw or defect in the weld, an importance of the weld, and a customer or project.
24 . A method of computer-assisted image analysis, comprising:
receiving a digital image from an imaging device;
performing a first determination as to determine whether a weld shown in the digital image conforms to weld specifications;
generating a digital ticket in response to the first determination, the digital ticket comprising data comprising a unique image identifier (ID) and a performance indicator representing a determination as to whether the digital image conforms to the weld specifications;
sending to a weld analytics system the digital image and the digital ticket;
receiving by the weld analytics system the digital image and the digital ticket;
storing by the weld analytics system the digital image in the repository;
performing by the weld analytics system a second determination as to determine whether a weld shown in the digital image conforms to weld specifications; and
generating and storing a weld record in the repository in response to the second determination, wherein the weld record comprises data comprising the unique image ID and the first performance indicator and/or a second performance indicator representing a determination as to whether the weld shown in the digital image conforms to the weld specifications.
25 . The method of claim 24 , further comprising:
determining whether the digital image conforms to image quality specifications.
26 . The method of claim 25 , wherein the first determination as to determine whether the weld shown in the digital image conforms to weld specifications is only performed in response to a determination that a digital image conforms to the image quality specifications.
27 . The method of claim 25 , further comprising:
generating an alert in response to a determination that the digital image does not conform to image specifications, wherein the alert is displayed on a display of the computer terminal.
28 . The method of claim 25 , wherein the image quality specifications specify a plurality of image qualities comprising one or more of an image resolution, an exposure level, a focus level, a contrast level, and a size of the weld in the digital image.
29 . The method of claim 25 , further comprising:
generate a digital ticket for each image that conforms to the image quality specifications, wherein the digital ticket comprises the first performance indicator and a unique ID to identify the respective image.
30 . The method of claim 29 , wherein the digital ticket further comprises one or more of a date the image was taken, a time the image was taken, a location at which the image was taken via telemetry, a welder who made the weld, a unique customer ID, and a unique job ID.
31 . The method of claim 29 , wherein the digital ticket further comprises, in response to a determination that the weld shown in the digital image does not conform to weld specifications, one or more of a location of a flaw or defect in the weld, a type of a flaw or defect in the weld, a description of the flaw or defect, a description of the weld in machine, and a likelihood that the weld is conforming.
32 . The method of claim 29 , wherein the digital ticket is generated in a human readable format and encoded with eXtensible Markup Language (XML) that can be extracted in downstream processing.
33 . The method of claim 24 , further comprising:
generating a notification in response to a determination that the weld shown in the digital image does not conform to weld specifications.
34 . The method of claim 33 , wherein the notification is an electronic message sent to one or more designated recipients, wherein the notification preferably includes information about the weld extracted from a weld record, and wherein the notification preferably includes a copy of the digital image of the weld and/or the corresponding digital ticket.
35 . The method of claim 24 , further comprising:
determining a likelihood that the weld shown in the image conforms to the weld specifications.
36 . The method of claim 24 , wherein the digital image is a real-time radiography (RTR) image.
37 . The method of claim 24 , wherein the imaging device is an ultrasonic imaging device and the digital image is an ultrasonic image.
38 . The method of claim 24 , wherein the imaging device is a computed radiography device or digital radiography imaging device and the digital image is a radiographic image.
39 . The method of claim 24 , wherein the weld analytics system has been trained by artificial intelligence (AI) or machine learning (ML).
40 . The method of claim 39 , wherein the weld analytics system comprises an image classifier trained to classify images of welds by the type of weld and/or whether an image of a weld shows a weld conforming to weld specifications corresponding to the type of weld.
41 . The method of claim 40 , wherein the image classifier comprises at least one neural network.
42 . The method of claim 24 , further comprising:
generating a new job record in the repository in response to a determination that the weld is non-conforming.
43 . The method of claim 42 , further comprising:
generating a work order for repairing or fixing the non-conforming weld, wherein the work order preferably includes information generated from the corresponding weld record, wherein the work order preferably includes a technician and/or welder to perform the work, and a date on which the work is to be performed; and
electronically sending the work order to one or more users.
44 . The method of claim 43 , wherein the work order comprises at least a scheduled date on which the work is to be performed, wherein the scheduled date is based on one or more of an importance of the weld, a customer and/or job, and construction due dates and/or timelines/schedule.
45 . The method of claim 44 , wherein the work order further comprises a technician and/or welder to perform the work, wherein the technician and/or welder is based on one or more of the welder who formed the non-conforming weld, a location of a flaw or defect in the weld, a type of flaw or defect in the weld, an importance of the weld, and a customer or project.
46 . A method of computer-assisted image analysis performed by a weld analytics system, comprising:
receiving a digital image from an imaging device and a digital ticket, the digital ticket comprising data comprising a unique image identifier (ID) and a performance indicator representing a determination as to whether the digital image conforms to the weld specifications;
storing the digital image in a repository;
performing a determination as to determine whether a weld shown in the digital image conforms to weld specifications; and
generating and storing a weld record in the repository in response to the second determination, wherein the weld record comprises data comprising the unique image ID and a second performance indicator representing a determination as to whether the weld shown in the digital image conforms to the weld specifications.