Real-time tracking method based on on-line learning and tracking system thereof

Abstract

The invention provides a real-time tracking method based on on-line learning and a tracking system thereof. The method comprises the following steps: acquiring image information through an image sensor, manually selecting an initial positive sample, utilizing a corresponding processing unit to complete extraction of object characteristic information, realizing classification of an object according to random forest, introducing classification reliability degree, according to a program similar to the object characteristic information, determining a positive sample and a negative sample which are used for training a random forest classifier, and finally utilizing a KLT tracking algorithm to realize accurate and high precision tracking of an object.

Description

technical field [0001] The invention relates to the field of video motion image data processing and machine vision, in particular to a real-time tracking method and tracking system based on online learning. Background technique [0002] Online learning belongs to the research category of incremental learning. In this type of method, the classifier only learns each sample once instead of repeated learning. In this way, a large amount of storage space is not required to store training samples during the operation of the online learning algorithm. Every time the classifier obtains a sample, it can be deleted after its learning is completed. [0003] Online learning greatly weakens the tedious step of manual labeling in the learning process. We only need to manually label a small sample set for the initial training of the classifier, and then the classifier can continuously obtain new samples when performing classification tasks. , so as to continuously self-train and improve t...

AI Technical Summary

Problems solved by technology

However, the integrated learning method is an offline learning method. The classifier trained by this method is fixed and cannot update and improve its own classification ability by learning new samples.

Once the training samples are not complete or the selected features are inappropriate, the entire classifier must be retrained and cannot be improved on the original basis, which will consume a lot of time and manpower, so the online learning method came into being

Images