Back to results
US 7,031,496 B2Granted
Method and apparatus for object recognition using a plurality of cameras and databases
AssigneeMatsushita Electric Industrial.Co., Ltd.
InventorsMihoko Shimano, Kenji Nagao, Toshiaki Akimoto, Tomonobu Naruoka
Life of this patent
Estimating the term…
Abstract
Based on image data taken and acquired by any one of a plurality of cameras, a predetermined object included in the image is recognized by pattern matching. A plurality of databases associated with a plurality of cameras is provided and a database to be used is changed according to the camera to be used. The plurality of databases can also be changed according to conditions such as weather and environment.
Claims9
An object recognition apparatus, comprising:
- a plurality of cameras that each take an image of an object and obtain image data of said object;
- a plurality of databases that each have registered therein a plurality of model data concerning object models, each of the plurality of databases being associated with one of the plurality of cameras and one of a condition of an environment and a combination of conditions of the environment;
- a search range focusing section that narrows down an area in each image data where a corresponding object is likely to exist and makes said area a search range;
- and
- an object recognition section that selects one of the plurality of databases based on a camera corresponding to the image data and one of a detected condition of the environment and a combination of detected conditions of the environment, compares the image data in the search range in each image data with the plurality of model data in the selected database, detects model data presenting a highest similarity to the image data, and detects the object in the image data using said model data,
- wherein the object recognition section is provided for common use by the plurality of cameras.
3 more elements
An object recognition apparatus, comprising:
- a plurality of cameras that each generate image data;
- a plurality of databases that store model object data, each of the plurality of databases being associated with one of the plurality of cameras and one of a condition of an environment and a combination of conditions of the environment;
- a search range focusing section that selects a reduced portion of the image data;
- and
- an object recognition section that selects one of the plurality of databases based on a camera corresponding to the image data and one of a detected condition of the environment and a combination of detected conditions of the environment, compares the selected portion of the image data to the stored model object data, selects stored model object data that has a highest similarity to the selected reduced portion of the image data, and detects an object from the image data using the selected model object data.
2 more elements
A method for recognizing an object, comprising:
- generating image data with one of a plurality of cameras;
- selecting a reduced portion of the image data;
- selecting a database from a plurality of databases based on a camera that generated the image data and one of a detected condition of an environment and a combination of detected conditions of the environment, each of the plurality of databases being associated with one of the plurality of cameras and one of a condition of the environment and a combination of conditions of the environment;
- comparing the selected portion of the image data to model object data stored in the selected database;
- selecting the stored model object data that has a highest similarity to the selected portion of the image data;
- and
- detecting an object from the image data using the selected stored model object data.
4 more elements
Dependencies are inferred from the claim text.