Object recognition method and recognition apparatus

Abstract

The invention provides an object recognition method and a recognition apparatus which can recognize objects by directly use a distorted fish-eye image for object recognition without projective conversion or image conversion based on corrective operation to the fish-eye image. A target fish-eye image (100) for object recognition is split (120) into regions in accordance with the direction of distortion of the target fish-eye image (100). An object recognition portion (130) performs object recognition by using databases (140) prepared in accordance with the regions respectively. The same process is applied to a plurality of target fish-eye images (110) created by rotation of the target fish-eye image (100). A detected coordinate transformation portion (150) restores the obtained position of an object into the original position by inverse rotation, and outputs a thus obtained result as a detection result. Accordingly, detection accuracy in object recognition can be improved, and the quality of data in databases can be reduced. Rotated object images required for database creation are used for creation of databases for object recognition for the fish-eye image.

Description

technical field [0001] The invention relates to an object recognition method and a recognition device, in particular to an object recognition method and a recognition device for recognizing a photographed object including a person in an image captured by a fisheye lens. Background technique [0002] The fisheye lens has the advantage of a wide field of view, but also has the disadvantage of including large distortion as the captured image moves from the center to the outside. Therefore, as a method for recognizing a captured object within an image captured by a fisheye lens, the prior art object recognition method used for a planar image with no or little distortion cannot be directly used. [0003] In order to be able to recognize a captured object in an image captured by a fisheye lens using the prior art object recognition method used for a planar image, it is generally known to convert a distorted image obtained by using a fisheye lens into a The technique of distorting...

AI Technical Summary

Problems solved by technology

However, in this case, it is necessary to generate a database covering all patterns of distortion appearing on the fisheye image, and the generation of the database requires more labor than the generation of the database using a flat image.

[0012] Also, in the case described above, the hardware resources required for the database storing data with various patterns of distortion are also larger than when the distortion is not considered

[0013] Furthermore, in the above case, if the image of the object in any certain area on the fisheye image is detected, the database that stores the image data with all modes of distortion is directly used, since the above-mentioned database also includes objects with The distortion of the area is not similar to the database of the distorted object data, so it is still not possible to detect the object with sufficient detection accuracy

Images