Abstract: A focus position detection device calculates, for each of shift amount calculation areas in a measurement area set on an image sensor of an imaging unit, a local shift amount of a second sub-image with respect to a first sub-image when a subject on the first sub-image and a subject on the second sub-image are most coincident with each other and a feature quantity of a probability distribution representing accuracy of the local shift amount, estimates the probability distribution based on the local shift amount and the feature quantity for each shift amount calculation area, combines the probability distributions of the shift amount calculation areas to obtain a distribution of accuracy of the local shift amount for the measurement area, and calculates a representative value representing the distance between a focus position produced by an optical system of the imaging unit and the image sensor based on the distribution.

Abstract: A method of estimating a phase difference includes: setting an area that is to be focused in an imaging range of an imaging device, the imaging device including an imaging element having a plurality of pixel arrays of phase-different pixels; first calculating, when a representative value for the plurality of pixel arrays is calculated, a pixel reference direction in which a pixel value is referred to, based on a position of an edge that appears in the plurality of pixel arrays; executing statistical processing of the pixel value for the plurality of pixel arrays in the calculated reference direction; and second calculating, by a processor, a phase difference using a pixel array that represents the plurality of pixel arrays that have been calculated by the statistical processing.

Abstract: A focus position detection apparatus calculates, for each shift amount calculation area in a measurement area set on an image sensor of a camera, a local shift amount between a first sub-image generated by first pixels and a second sub-image generated by second pixels, and a degree of reliability of the local shift amount, and corrects the degree of reliability base on a level of inclusion of components included in a subject image in the shift amount calculation area and being finer than a distance between two adjacent first pixels or a distance between two adjacent second pixels. The focus position detection apparatus then calculates a representative value representing a distance between a focus position by an optical system of the camera and the image sensor so that influence of the local shift amount of each shift amount calculation area increases as the corrected degree of reliability increases.

Abstract: A focus position detection device calculates, for each shift amount calculation area contained in a measurement area defined on an image sensor, a local shift amount representing a shift between a first sub-image generated from first pixels and a second sub-image generated from second pixels and its confidence score, and corrects the confidence score based on at least one of the spacing between the first pixels, the spacing between the second pixels, and the amount of positional displacement between the first and second pixels in a direction orthogonal to the edge direction of a subject captured in the shift amount calculation area. Then, the focus position detection device calculates a representative value representing the distance between the image sensor and the focus position, by taking a weighted average of the local shift amounts of the respective shift amount calculation areas with weighting based on the corrected confidence scores.

Abstract: A focus position detection device calculates, for each of shift amount calculation areas in a measurement area set on an image sensor of an imaging unit, a local shift amount of a second sub-image with respect to a first sub-image when a subject on the first sub-image and a subject on the second sub-image are most coincident with each other and a feature quantity of a probability distribution representing accuracy of the local shift amount, estimates the probability distribution based on the local shift amount and the feature quantity for each shift amount calculation area, combines the probability distributions of the shift amount calculation areas to obtain a distribution of accuracy of the local shift amount for the measurement area, and calculates a representative value representing the distance between a focus position produced by an optical system of the imaging unit and the image sensor based on the distribution.

Abstract: A processor of executes the following processes on one or more second feature point pairs for a first feature point pair from among created first pair candidates. A first vector is a vector that connects a first feature point of the first feature point pair and a first feature point of the second feature point pair. A second vector is a vector that connects a second feature point of the first feature point pair and a second feature point of the second feature point pair. The processor determines that a first feature point pair has been deleted from the first pair candidates, on the basis of a comparison result between a predetermined angle and an angle between the first vector and the second vector. The processor performs matching between the first image and the second image on the basis of feature point.

Abstract: A processor of executes the following processes on one or more second feature point pairs for a first feature point pair from among created first pair candidates. A first vector is a vector that connects a first feature point of the first feature point pair and a first feature point of the second feature point pair. A second vector is a vector that connects a second feature point of the first feature point pair and a second feature point of the second feature point pair. The processor determines that a first feature point pair has been deleted from the first pair candidates, on the basis of a comparison result between a predetermined angle and an angle between the first vector and the second vector. The processor performs matching between the first image and the second image on the basis of feature point.