Image recognition device, method for image recognition device, and recording medium
It is necessary to input and register the size of an object in advance for each installation environment, which takes time and effort. This image recognition device comprises: an image acquisition unit ( 11 ) that acquires an image; a detection unit ( 12 ) that detects a detection object from the image; an index creation unit ( 14 ) that creates an index for determining whether the detection object is a target object from the history of the location and size of the detection object; and a determination unit ( 15 ) that uses the index to determine whether the detection object is the target object, wherein the index is information representing a distribution trend of the location and size of the detection object in the history.
1 . An image recognition device, comprising:
a processor configured to:
acquire a captured image;
detect a predetermined type of object as a detection object by image recognition from the captured image;
perform control to record a history of a location and a size of the detection object detected from the captured image;
create an index for determining whether the detection object is a target object to be detected based on the history; and
determine whether the detection object is the target object by comparing the location and the size of the detection object detected from the captured image with the index, wherein
the processor creates a function indicating a relationship between the location of the detection object and the size of the detection object as the index and determines that the detection object is the target object in response to a distance determined from the function and the location and the size of the detection object in the captured image being less than a predetermined threshold value, wherein
the processor performs control to record a correspondence relationship between a one-dimensional location of the detection object and a size of the detection object in a predetermined uniaxial direction in the captured image, the function is a two-dimensional approximate curve between the one-dimensional location of the detection object and the size of the detection object, and the distance is a distance between a point representing the one-dimensional location of the detection object and the size of the detection object and the approximate curve; or
the processor performs control to record a correspondence relationship between a two-dimensional location of the detection object and a size of the detection object in predetermined two axial directions in the captured image, the function is a three-dimensional approximate curved surface of the two-dimensional location of the detection object and the size of the detection object, and the distance is a distance between a point representing the two-dimensional location of the detection object and the size of the detection object and the approximate curved surface.
2 . A method for an image recognition device, comprising:
an image acquisition step of acquiring a captured image;
a detection step of detecting a predetermined type of object as a detection object by image recognition from the captured image;
a recording control step of performing control to record a history of a location and a size of the detection object in the captured image detected in the detection step;
an index creation step of creating an index for determining whether the detection object is a target object to be detected based on the history;
a determining step of determining whether the detection object is the target object by comparing the location and the size of the detection object in the captured image detected in the detection step with the index, wherein
a function indicating a relationship between the location of the detection object and the size of the detection object is created as the index and the detection object is determined as the target object in response to a distance determined from the function and the location and the size of the detection object in the captured image being less than a predetermined threshold value; and
a step of performing control to record a correspondence relationship between a one-dimensional location of the detection object and a size of the detection object in a predetermined uniaxial direction in the captured image, wherein the function is a two-dimensional approximate curve between the one-dimensional location of the detection object and the size of the detection object, and the distance is a distance between a point representing the one-dimensional location of the detection object and the size of the detection object and the approximate curve; or
a step of performing control to record a correspondence relationship between a two-dimensional location of the detection object and a size of the detection object in predetermined two axial directions in the captured image, wherein the function is a three-dimensional approximate curved surface of the two-dimensional location of the detection object and the size of the detection object, and the distance is a distance between a point representing the two-dimensional location of the detection object and the size of the detection object and the approximate curved surface.
3 . A non-transient computer-readable recording medium, recording a program for an image recognition device, being an image recognition program for causing a computer to function as the image recognition device according to claim 1 , and the program being configured to cause a computer to function as the processor.