Detection device, detection system, detection method, and storage medium

A detection method, comprising:

• detecting an object from a first viewpoint;
• calculating first model information including shape information on the object from the first viewpoint by using detection results in the detecting;
• calculating, by using a first taken image obtained by imaging a space including a light source that irradiates the object with illumination light and including the object, light source information on the light source, wherein the space has at least a hemispherical range centered at the first viewpoint;
• calculating a positional relation between the first viewpoint and the object by using the light source information as information used to integrate the first model information and second model information including shape information obtained by detecting the object from a second viewpoint different from the first viewpoint, wherein
• the light source information includes information on at least one of (i) the position of the light source that irradiates the object with the illumination light with reference to the object, (ii) an irradiation direction in which the light source irradiates the object with the illumination light, (iii) a wavelength of the illumination light emitted from the light source, and (iv) a type of the light source;
• and
• integrating the first model information and the second model information, to correlate the positional relation between the first viewpoint and the object and a positional relation between the second viewpoint and the object, by converting a coordinate system of the second model information into a coordinate system of the first model information and then adding the coordinate-converted second model information to the first model information.

A non-transitory computer-readable storage medium storing therein a detection program causing a computer to execute:

• calculating, by using detection results of detecting an object from a first viewpoint, first model information including shape information on the object from the first viewpoint;
• calculating, by using a first taken image obtained by imaging a space including a light source that irradiates the object with illumination light and including the object, light source information on the light source, wherein the space has at least a hemispherical range centered at the first viewpoint;
• calculating a positional relation between the first viewpoint and the object by using the light source information as information used to integrate the first model information and second model information including shape information obtained by detecting the object from a second viewpoint different from the first viewpoint, wherein
• the light source information includes information on at least one of (i) the position of the light source that irradiates the object with the illumination light with reference to the object, (ii) an irradiation direction in which the light source irradiates the object with the illumination light, (iii) a wavelength of the illumination light emitted from the light source, and (iv) a type of the light source;
• and
• integrating the first model information and the second model information, to correlate the positional relation between the first viewpoint and the object and a positional relation between the second viewpoint and the object, by converting a coordinate system of the second model information into a coordinate system of the first model information and then adding the coordinate-converted second model information to the first model information.

A detection device, comprising:

• a detector that detects an object from a first viewpoint;
• an information calculator that calculates first model information including shape information on the object from the first viewpoint by using detection results of the detector;
• a light source calculator that calculates light source information on a light source that irradiates the object with illumination light;
• a position calculator that calculates a positional relation between the first viewpoint and the object or the light source and a positional relation between a second viewpoint different from the first viewpoint, and the object or the light source, or a positional relation between the first viewpoint and the second viewpoint, by using the light source information as information used to integrate the first model information and second model information including shape information obtained by detecting the object from the second viewpoint;
• and
• a model integrator that integrates the first model information and the second model information by converting a coordinate system of the second model information into a coordinate system of the first model information and then adding the coordinate-converted second model information to the first model information.

A detection device, comprising:

• a first detector that detects an object from a first viewpoint;
• a second detector that detects the object from a second viewpoint;
• an information calculator that calculates first model information including shape information on the object from the first viewpoint and second model information including shape information on the object from the second viewpoint by using detection results of the first detector and the second detector;
• a light source calculator that calculates first light source information on a first light source that irradiates the object with illumination light, and that calculates second light source information on a second light source that irradiates the object with illumination light;
• a position calculator that calculates a first relation between the first light source and the first viewpoint and a second relation between the second light source and the second viewpoint;
• and
• a model integrator that integrates the first model information and the second model information, to correlate the first light source information and the second light source information, by converting a coordinate system of the second model information into a coordinate system of the first model information and then adding the coordinate-converted second model information to the first model information.