Method and apparatus for configuring and testing a machine vision detector
First Claim
1. A method for testing a vision detector comprising:
- continuously moving a sequence of objects relative to a field of view of the vision detector;
capturing a plurality of groups of frames, each frame of the plurality of groups of frames comprising an image of the field of view, such that each group of frames of the plurality of groups of frames comprises a plurality of images of one corresponding object in the sequence of objects;
analyzing each object to determine a plurality of results corresponding to the plurality of groups of frames, each result in the plurality of results comprising information about the status of the object corresponding to the group of frames;
storing, responsive to the plurality of results, a selected set of groups of frames from the plurality of groups of frames, the set of groups of frames comprising a stored set of images; and
displaying a portion of the stored set of images for viewing by a human user, the portion displayed chosen responsive to scrolling commands from the human user, the scrolling commands capable of advancing the portion displayed forward and backward in the stored set of images, whereineach result in the plurality of results comprises a number that indicates a relative confidence that the corresponding object was detected;
the number is compared to a decision threshold to decide whether the corresponding object was detected; and
the selected set of groups of frames corresponds to objects for which the corresponding numbers are close to the decision threshold.
3 Assignments
0 Petitions
Accused Products
Abstract
Disclosed are systems and methods for configuring a vision detector, wherein a training image is obtained from a production line operating in continuous motion so as to provide conditions substantially identical to those that will apply during actual manufacturing and inspection of objects. A training image can be obtained without any need for a trigger signal, whether or not the vision detector might use such a signal for inspecting the objects. Further disclosed are systems and methods for testing a vision detector by selecting, storing, and displaying a limited number of images from a production run, where those images correspond to objects likely to represent incorrect decisions.
206 Citations
12 Claims
-
1. A method for testing a vision detector comprising:
-
continuously moving a sequence of objects relative to a field of view of the vision detector; capturing a plurality of groups of frames, each frame of the plurality of groups of frames comprising an image of the field of view, such that each group of frames of the plurality of groups of frames comprises a plurality of images of one corresponding object in the sequence of objects; analyzing each object to determine a plurality of results corresponding to the plurality of groups of frames, each result in the plurality of results comprising information about the status of the object corresponding to the group of frames; storing, responsive to the plurality of results, a selected set of groups of frames from the plurality of groups of frames, the set of groups of frames comprising a stored set of images; and displaying a portion of the stored set of images for viewing by a human user, the portion displayed chosen responsive to scrolling commands from the human user, the scrolling commands capable of advancing the portion displayed forward and backward in the stored set of images, wherein each result in the plurality of results comprises a number that indicates a relative confidence that the corresponding object was detected; the number is compared to a decision threshold to decide whether the corresponding object was detected; and the selected set of groups of frames corresponds to objects for which the corresponding numbers are close to the decision threshold. - View Dependent Claims (2, 3, 4, 5, 6)
-
-
7. An apparatus comprising:
-
a sequence of objects; a vision detector having a field of view, the sequence of objects in continuous motion relative to the field of view, the vision detector configured to capture a plurality of groups of frames, each frame of the plurality of groups of frames comprising an image of the field of view, such that each group of frames of the plurality of groups of frames comprises a plurality of images of one corresponding object in the sequence of objects, analyze each object to determine a plurality of results corresponding to the plurality of groups of frames, each result in the plurality of results comprising information about the status of the object corresponding to the group of frames; a memory configured to store, responsive to the plurality of results, a selected set of groups of frames from the plurality of groups of frames, the set of groups of frames comprising a stored set of images; and a human-machine interface configured to display a portion of the stored set of images for viewing by a human user, the portion displayed chosen responsive to scrolling commands from the human user, the scrolling commands capable of advancing the portion displayed forward and backward in the stored set of images, wherein each result in the plurality of results comprises a number that indicates a relative confidence that the corresponding object was detected; the number is compared to a decision threshold to decide whether the corresponding object was detected; and the selected set of groups of frames corresponds to objects for which the corresponding numbers are close to the decision threshold. - View Dependent Claims (8, 9, 10, 11, 12)
-
Specification