Apparatus and method for tracking multiple objects
AssigneePostech Research and Business Development Foundation
InventorsDai Jin Kim, Hye Min Lee
Life of this patent
Estimating the term…
Abstract
Disclosed is a method of tracking multiple objects included in an image using a multi-object tracking apparatus including an integrated similarity neural network, the method including setting a tracking area in an input image, extracting at least one object candidate for a target object from the tracking area; extracting reference features for the target object, the object candidate, and the tracking area, selecting two of the target object, the object candidate, and the tracking area to evaluate similarity based on the reference features; allocating the object candidate to the target object on the basis of the evaluated similarity; and tracking the target object on the basis of a location of the allocated object candidate.
Claims16
A method of tracking multiple objects included in an image using a multi-object tracking apparatus including an integrated similarity neural network, the method comprising:
- setting a tracking area in an input image;
- extracting at least one object candidate for a target object from the tracking area;
- extracting reference features for the target object, the object candidate, and the tracking area;
- evaluating similarity between the target object and the object candidate by calculating a similarity value based on the reference features using the integrated similarity neural network;
- evaluating similarity between the target object and the tracking area by calculating a similarity map and a phase value of an object based on the reference features using the integrated similarity neural network;
- allocating the object candidate to the target object on the basis of the evaluated similarity;
- and
- tracking the target object on the basis of a location of the allocated object candidate,
- wherein both of the evaluating similarity between the target object and the object candidate and the evaluating similarity between the target object and the tracking area are being performed together in the integrated similarity neural network.
A multi-object tracking apparatus, comprising:
- a processor;
- and
- a memory configured to store at least one command executable by the processor,
- wherein the command comprises:
- a command to set a tracking area in an input image;
- a command to extract at least one object candidate for a target object from the tracking area;
- a command to extract reference features for the target object, the object candidate, and the tracking area;
- a command to evaluate similarity between the target object and the object candidate by calculating a similarity value based on the reference features using an integrated similarity neural network;
- a command to evaluate similarity between the target object and the tracking area by calculating a similarity map and a phase value of an object based on the reference features using the integrated similarity neural network;
- a command to allocate the object candidate to the target object on the basis of the evaluated similarity;
- and
- a command to track the target object on the basis of a location of the allocated object candidate, and
- wherein both of the command to evaluate similarity between the target object and the object candidate and the command to evaluate similarity between the target object and the tracking area are for being performed together in the integrated similarity neural network.
Dependencies are inferred from the claim text.