主权项 |
1. A method for image recognition using a camera system, the camera system including a plurality of camera devices, the method comprising:
acquiring, by a communication interface and from a first camera device of the plurality of camera devices, a first image and first imaging environment information, wherein the first image is taken by the first camera device, and the first imaging environment information indicates a first imaging environment of the first camera device at a time of taking the first image; determining, by a processor using a parameter table that stores previous imaging environment information indicating imaging environments of previous images taken at previous times by the plurality of camera devices and recognition control parameters corresponding to the imaging environments, a first recognition control parameter based on the first imaging environment information, wherein the first recognition control parameter indicates a first detector corresponding to a second imaging environment of a second image taken by a second camera device of the plurality of camera devices, the second imaging environment having a similarity that is identical or of a degree that exceeds a predetermined threshold with respect to the first imaging environment; and performing, by the processor, image recognition of the first image acquired from the first camera device, using the first detector indicated by the first recognition control parameter, wherein the previous imaging environment information indicating the imaging environments of the previous images taken at the previous times by the plurality of camera devices includes information on any of surrounding illuminance of the previous images, illumination orientation of the plurality of camera devices, distortion due to lenses attached to the plurality of camera devices, shake of the previous images due to shake of the plurality of camera devices, target object size in the previous images, and target object orientation in the previous images. |