Unlock instant, AI-driven research and patent intelligence for your innovation.
A target tracking method and device based on adaptive search
What is Al technical title?
Al technical title is built by PatSnap Al team. It summarizes the technical point description of the patent document.
An adaptive and position tracking technology, applied in the field of computer vision, can solve problems such as tracking result deviation, inability to track target objects, tracking performance impact, etc.
Active Publication Date: 2021-02-05
BOCOM SMART INFORMATION TECH CO LTD
View PDF6 Cites 0 Cited by
Summary
Abstract
Description
Claims
Application Information
AI Technical Summary
This helps you quickly interpret patents by identifying the three key elements:
Problems solved by technology
Method used
Benefits of technology
Problems solved by technology
[0004] However, there are some inherent problems in the kernel correlation filter tracking algorithm: first, when the target object is continuously occluded, the occluder will block the target, which will cause the tracking result to deviate
Again, if the target object moves fast and exceeds the search area of the tracking algorithm, the target object cannot be tracked, and the tracking performance will also be affected
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more
Image
Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
Click on the blue label to locate the original text in one second.
Reading with bidirectional positioning of images and text.
Smart Image
Examples
Experimental program
Comparison scheme
Effect test
Embodiment 1
[0037] This embodiment provides an adaptive search-based target tracking method for tracking target objects, especially fast-moving target objects can be better tracked. The method can track a target object in an input video in smart devices such as a user's computer and a server.
[0038] The target tracking method based on adaptive search in this embodiment is as follows figure 1 shown, including the following steps:
[0039] S1. Obtain the original position of the target object. At the initial frame of the video, the initial position of the target is given.
[0040] S2. Track the target object according to the original position to obtain a tracking position, which is a rectangular frame. In the step of tracking the target object, a kernel correlation filter tracking algorithm is used for tracking.
[0041] The tracking process of the kernel correlation filtering algorithm can be decomposed into the following steps:
[0042] In the first step, in the initial frame image...
Embodiment 2
[0078] In this embodiment, a target tracking device based on adaptive search is provided, which is used to track target objects in video images, and is especially suitable for fast-moving targets and occluded targets. The structure of the device is as follows figure 2 shown, including:
[0079] The original position extraction unit 21 is used to obtain the original position of the target object;
[0080] The tracking unit 22 is configured to track the target object according to the original position to obtain a tracked position; and use a kernel correlation filter tracking algorithm for tracking.
[0081] A similarity calculation unit 23, configured to calculate the similarity between the tracking position and the original position of the target object;
[0082] A judging unit 24, configured to judge whether the similarity is less than a preset threshold;
[0083] The first tracking unit 25 is configured to obtain a plurality of candidate tracking regions through a cascade ...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More
PUM
Login to View More
Abstract
The target tracking method based on adaptive search provided by the present invention first tracks the target object to obtain the tracking position, and then calculates the similarity between the tracking position and the original position of the target object, if the similarity is less than the preset threshold , obtain multiple candidate tracking regions through cascaded detectors, and cluster the candidate tracking regions with the highest confidence as the tracking result; if the similarity is not less than the preset threshold, then obtain the target object in the tracking position Size, update the original target object according to the size of the target object, and track the target object according to the updated target size in the next frame of image. In this scheme, if the difference from the original position is too large, it means that the target moves rapidly and the tracking fails, and if the difference is small, it means the tracking is successful. This method is suitable for tracking when the target object is occluded or the target object moves fast, and has better tracking effect.
Description
technical field [0001] The invention relates to the field of computer vision, in particular to an adaptive search-based target tracking method and device. Background technique [0002] In the field of computer vision, object tracking has always been one of the hot research areas. The so-called target tracking is the process of continuously locating the target of interest in a continuous image sequence. Target tracking is widely used in military, transportation, monitoring and other fields. Due to factors such as illumination changes, target deformation, target occlusion, and real-time performance, it is still difficult to achieve accurate target tracking algorithms. [0003] Kernel correlation filter tracking algorithm is a tracking algorithm with better effect developed in recent years. It uses the idea of cyclic shift to construct a large number of samples to train the classifier, and uses discrete Fourier transform to reduce the training and detection process of the c...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More
Application Information
Patent Timeline
Application Date:The date an application was filed.
Publication Date:The date a patent or application was officially published.
First Publication Date:The earliest publication date of a patent with the same application number.
Issue Date:Publication date of the patent grant document.
PCT Entry Date:The Entry date of PCT National Phase.
Estimated Expiry Date:The statutory expiry date of a patent right according to the Patent Law, and it is the longest term of protection that the patent right can achieve without the termination of the patent right due to other reasons(Term extension factor has been taken into account ).
Invalid Date:Actual expiry date is based on effective date or publication date of legal transaction data of invalid patent.
Login to View More 
Login to View More