20240029394. OBJECT DETECTION METHOD, OBJECT DETECTION DEVICE, AND PROGRAM simplified abstract (Konica Minolta, Inc.)
Contents
OBJECT DETECTION METHOD, OBJECT DETECTION DEVICE, AND PROGRAM
Organization Name
Inventor(s)
OBJECT DETECTION METHOD, OBJECT DETECTION DEVICE, AND PROGRAM - A simplified explanation of the abstract
This abstract first appeared for US patent application 20240029394 titled 'OBJECT DETECTION METHOD, OBJECT DETECTION DEVICE, AND PROGRAM
Simplified Explanation
The patent application describes an object detection method that involves estimating key point candidates for each object in an image and detecting key points based on these estimated candidates. The key points are points that satisfy a defined condition among points indicating the boundary of the object model projected onto defined coordinate axes. These coordinate axes have an origin at the geometric center of the object model and form a defined angle relative to a polar axis in a polar coordinate system set for the object model.
- Key point estimation step: Estimating key point candidates for each object in an image.
- Detection step: Detecting key points for each object based on the estimated key point candidates.
- Object model: Models the shape of an object.
- Key points: Points that satisfy a defined condition among points indicating the boundary of the object model projected onto defined coordinate axes.
- Coordinate axes: Have an origin at the geometric center of the object model and form a defined angle relative to a polar axis in a polar coordinate system set for the object model.
Potential Applications:
- Object detection and recognition in computer vision systems.
- Autonomous vehicles for detecting and tracking objects.
- Robotics for object manipulation and interaction.
Problems Solved:
- Accurate detection and localization of objects in images.
- Efficient estimation of key points for object recognition.
- Handling complex object shapes and variations.
Benefits:
- Improved object detection accuracy.
- Robustness to object shape variations.
- Efficient and reliable key point estimation.
- Applicable to various domains such as computer vision, robotics, and autonomous systems.
Original Abstract Submitted
an object detection method includes a key point estimation step of estimating key point candidates for each object in an image; and a detection step of detecting key points for each object based on the estimated key point candidates. considering an object model that models shape of an object, the key points are points that satisfy a defined condition among points indicating a boundary of the object model that are projected onto defined coordinate axes. the defined coordinate axes have an origin at a geometric center of the object model and each forms a defined angle relative to a polar axis in a polar coordinate system set for the object model.
