An Improved Algorithm on Adaptive KLT Vision Tracking

2013 ◽  
Vol 631-632 ◽  
pp. 1270-1275
Author(s):  
Yuan Min Liu ◽  
Lian Fang Tian
Keyword(s):  
State Of The Art ◽  
Tracking System ◽  
Tracking Algorithm ◽  
Experimental Results ◽  
Feature Points ◽  
Backward Error ◽  
Tracking Method ◽  
Adaptive Tracking ◽  
Conventional Methods ◽  
Measurement Validity

In view of the shortage of the KLT (Kanade-Lucas-Tomasi) tracking algorithm, an improved adaptive tracking method based on KLT is proposed in this paper, in which a kind of filtering mechanism is applied to decrease the effects of noise and illumination on tracking system. In order to eliminate the error of tracking, the strategies based on forward-backward error and measurement validity are utilized properly. However, because the approach to forward-backward error makes the feature points reduce, which leads to tracking failure especially when the shapes of object change, a method for appending the feature points is introduced. Experimental results indicate that the method presented in this paper is state of the art robustness in our comparison with related work and demonstrate improved generalization over the conventional methods.

An Adaptive KLT Algorithm for Hand Gesture Tracking

2013 ◽  
Vol 706-708 ◽  
pp. 623-628
Author(s):  
Huang Xin
Keyword(s):  
Human Computer Interaction ◽  
State Of The Art ◽  
Tracking System ◽  
Adaptive Method ◽  
Experimental Results ◽  
Hand Gesture ◽  
Feature Points ◽  
Tracking Algorithms ◽  
Gesture Tracking ◽  
Conventional Methods

With the development of human-computer interaction technology, hand gesture is widely investigated recently for its natural and convenient properties. In view of the disadvantage of the existing tracking algorithms for the hand gesture, a novel adaptive method based on KLT is proposed in this paper, in which a kind of filtering mechanism is applied to decrease the effects of noise and illumination on tracking system. In order to eliminate the error of tracking, the strategy based on confidence is utilized properly. However, because the hand is non-rigid, its shape often changes, which easily leads to tracking failure for the reduction of features. In order to solve the problem, a method for appending the feature points is introduced. Experimental results indicate that the method presented in this paper is state of the art robustness in our comparison with related work and demonstrate improved generalization over the conventional methods.

scholarly journals Long-term target tracking combined with re-detection

2020 ◽  
Author(s):  
Juanjuan Wang ◽  
HaoRan Yang ◽  
Ning Xu ◽  
Chengqin Wu ◽  
ZengShun Zhao ◽  
...  
Keyword(s):  
Correlation Energy ◽  
State Of The Art ◽  
Tracking Algorithm ◽  
Correlation Filters ◽  
Short Term ◽  
Tracking Method ◽  
Tracking Tasks ◽  
Svm Model ◽  
Confidence Degree

Abstract The long-term visual tracking undergoes more challenges and is closer to realistic applications than short-term tracking. However, the performances of most existing methods have been limited in the long-term tracking tasks. In this work, we present a reliable yet simple long-term tracking method, which extends the state-of-the-art Learning Adaptive Discriminative Correlation Filters (LADCF) tracking algorithm with a re-detection component based on the SVM model. The LADCF tracking algorithm localizes the target in each frame and the re-detector is able to efficiently re-detect the target in the whole image when the tracking fails. We further introduce a robust confidence degree evaluation criterion that combines the maximum response criterion and the average peak-to correlation energy (APCE) to judge the confidence level of the predicted target. When the confidence degree is generally high, the SVM is updated accordingly. If the confidence drops sharply, the SVM re-detects the target. We perform extensive experiments on the OTB-2015 and UAV123 datasets. The experimental results demonstrate the effectiveness of our algorithm in long-term tracking.

scholarly journals Part-Based Visual Tracking via Online Weighted P-N Learning

2014 ◽  
Vol 2014 ◽  
pp. 1-13 ◽  
Author(s):  
Heng Fan ◽  
Jinhai Xiang ◽  
Jun Xu ◽  
Honghong Liao
Keyword(s):  
Visual Tracking ◽  
State Of The Art ◽  
The State ◽  
Tracking Algorithm ◽  
Local Feature ◽  
Experimental Results ◽  
Learning Method ◽  
Target Model ◽  
Tracking Process

We propose a novel part-based tracking algorithm using online weighted P-N learning. An online weighted P-N learning method is implemented via considering the weight of samples during classification, which improves the performance of classifier. We apply weighted P-N learning to track a part-based target model instead of whole target. In doing so, object is segmented into fragments and parts of them are selected as local feature blocks (LFBs). Then, the weighted P-N learning is employed to train classifier for each local feature block (LFB). Each LFB is tracked through the corresponding classifier, respectively. According to the tracking results of LFBs, object can be then located. During tracking process, to solve the issues of occlusion or pose change, we use a substitute strategy to dynamically update the set of LFB, which makes our tracker robust. Experimental results demonstrate that the proposed method outperforms the state-of-the-art trackers.

A New Algorithm of Image Matching Combining Sift and Shape Context

2014 ◽  
Vol 644-650 ◽  
pp. 4307-4313
Author(s):  
Yan Lu Xu ◽  
Yan Ma ◽  
Shun Bao Li ◽  
Ning Li Zhang ◽  
Xiang Fen Zhang ◽  
...  
Keyword(s):  
Image Matching ◽  
Experimental Results ◽  
Feature Points ◽  
Shape Context ◽  
Matching Accuracy ◽  
Conventional Methods ◽  
Distance Formula

This paper presents a new algorithm of image matching via combining sift and shape context for improving image matching accuracy. A joint descriptor is applied to describe feature points. The initial matching is obtained by a proposed distance formula. Furthermore, PLS is introduced to eliminate mismatched points. Experimental results demonstrate the proposed algorithm can achieve better performance compared to conventional methods.

Multi-Object Tracking with Single Camera

2015 ◽  
Vol 740 ◽  
pp. 668-671
Author(s):  
Yu Bing Dong ◽  
Ying Sun ◽  
Ming Jing Li
Keyword(s):  
Computer Vision ◽  
Object Tracking ◽  
Block Matching ◽  
Tracking Algorithm ◽  
Experimental Results ◽  
Single Camera ◽  
Trajectory Prediction ◽  
Tracking Method

Multi-object tracking has been a challenging topic in computer vision. A Simple and efficient moving multi-object tracking algorithm is proposed. A new tracking method combined with trajectory prediction and a sub-block matching is used to handle the objects occlusion. The experimental results show that the proposed algorithm has good performance.

Variable Scale Compressive Tracking Based on Structural Constraint Sample

2014 ◽  
Vol 488-489 ◽  
pp. 1074-1078
Author(s):  
Lu Ping Zhang ◽  
Meng Cai ◽  
Biao Li ◽  
Lu Ping Wang
Keyword(s):  
State Of The Art ◽  
Tracking Algorithm ◽  
Experimental Results ◽  
Structural Constraint ◽  
Scale Problem ◽  
Scale Selection ◽  
Compressive Tracking ◽  
Image Patches ◽  
Basic Attitude ◽  
Selection Of

A variable scale compressive tracking algorithm based on structural constraint sample is presented to solve the variable scale problem in this paper. A number of scanning windows with different scales and positions are obtained by structural constraint sampling.Some sparse random sensing matrices with different scales that can be computed offline easily are adopted to extract the features of different foreground target and background sample image patches with relevant scales online, the sample patch having a maximal score is regarded as the new tracking result by classifying the compressive features via a naive bayesian classifier,meanwhile,to update the location and scale. Experimental results show the proposed algorithm performs favorably against state-of-the-art algorithms on challenging sequences in terms of the basic attitude and scale change, which is robust and does not depend on the scale selection of the initial tracking area.

Adaptive Tracking Algorithm of Multi-Target Based on Fuzzy Clustering

2014 ◽  
Vol 513-517 ◽  
pp. 448-452
Author(s):  
Xiu Hua Hu ◽  
Lei Guo ◽  
Hui Hui Li
Keyword(s):  
State Estimation ◽  
Target Tracking ◽  
Fuzzy Clustering ◽  
Tracking System ◽  
Joint Probability ◽  
Tracking Algorithm ◽  
Adaptive Tracking ◽  
Correlation Criterion ◽  
Multi Target Tracking ◽  
Fuzzy Clustering Theory

For multi-target tracking system, aiming at solving the problem of low precision of state estimation caused by the data correlation ambiguity, the paper presents a novel multi-sensor multi-target adaptive tracking algorithm based on fuzzy clustering theory. Based on the joint probability data association algorithm, the new approach takes account of the case that whether the measure is validated and its possibility of belong to false alarm, and improves the correlation criterion of effective measurement with existing track on the basis of fuzzy clustering theory, which all perfect the update equation of target state estimation and the covariance. Meanwhile, with the adaptive distributed fusion processing structure, it enhance the robustness of the system and without prejudice to the real-time tracking. With the simulation case studies of radar/infrared sensor fusion multi-target tracking system, it verifies the effectiveness of the proposed approach.

scholarly journals Cross-Modality Paired-Images Generation for RGB-Infrared Person Re-Identification

2020 ◽  
Vol 34 (07) ◽  
pp. 12144-12151
Author(s):  
Guan-An Wang ◽  
Tianzhu Zhang ◽  
Yang Yang ◽  
Jian Cheng ◽  
Jianlong Chang ◽  
...  
Keyword(s):  
State Of The Art ◽  
Experimental Results ◽  
Fine Grained ◽  
Invariant Features ◽  
Proposed Model ◽  
Art Methods ◽  
Conventional Methods

RGB-Infrared (IR) person re-identification is very challenging due to the large cross-modality variations between RGB and IR images. The key solution is to learn aligned features to the bridge RGB and IR modalities. However, due to the lack of correspondence labels between every pair of RGB and IR images, most methods try to alleviate the variations with set-level alignment by reducing the distance between the entire RGB and IR sets. However, this set-level alignment may lead to misalignment of some instances, which limits the performance for RGB-IR Re-ID. Different from existing methods, in this paper, we propose to generate cross-modality paired-images and perform both global set-level and fine-grained instance-level alignments. Our proposed method enjoys several merits. First, our method can perform set-level alignment by disentangling modality-specific and modality-invariant features. Compared with conventional methods, ours can explicitly remove the modality-specific features and the modality variation can be better reduced. Second, given cross-modality unpaired-images of a person, our method can generate cross-modality paired images from exchanged images. With them, we can directly perform instance-level alignment by minimizing distances of every pair of images. Extensive experimental results on two standard benchmarks demonstrate that the proposed model favourably against state-of-the-art methods. Especially, on SYSU-MM01 dataset, our model can achieve a gain of 9.2% and 7.7% in terms of Rank-1 and mAP. Code is available at https://github.com/wangguanan/JSIA-ReID.

scholarly journals Visual Tracking of Small Unmanned Aerial Vehicles Based on Object Proposal Voting

2021 ◽  
Vol 11 (3) ◽  
pp. 953
Author(s):  
Jin Hong ◽  
Junseok Kwon
Keyword(s):  
Unmanned Aerial Vehicles ◽  
Visual Tracking ◽  
State Of The Art ◽  
Experimental Results ◽  
Tracking Method ◽  
Aerial Vehicles ◽  
Object Proposal ◽  
Background Clutter ◽  
Art Performance ◽  
Over Time

In this paper, we propose a novel visual tracking method for unmanned aerial vehicles (UAVs) in aerial scenery. To track the UAVs robustly, we present a new object proposal method that can accurately determine the object regions that are likely to exist. The proposed object proposal method is robust to small objects and severe background clutter. For this, we vote on candidate areas of the object and increase or decrease the weight of the area accordingly. Thus, the method can accurately propose the object areas that can be used to track small-sized UAVs with the assumption that their motion is smooth over time. Experimental results verify that UAVs are accurately tracked even when they are very small and the background is complex. The proposed method qualitatively and quantitatively delivers state-of-the-art performance in comparison with conventional object proposal-based methods.

The Target Tracking Method Based on Camshift Algorithm Combined with SIFT

2011 ◽  
Vol 186 ◽  
pp. 281-286 ◽  
Author(s):  
Jie Yu Zhang ◽  
Hai Yong Wu ◽  
Shu Chen ◽  
De Shen Xia
Keyword(s):  
Probability Distribution ◽  
Target Tracking ◽  
Experimental Results ◽  
Feature Points ◽  
Robust Tracking ◽  
Color Information ◽  
Search Region ◽  
Target Region ◽  
Tracking Method ◽  
Camshift Algorithm

Since Camshift algorithm leads to failed tracking results when the color information of the target region is similar with the background or is not precise enough, to solve this problem a tracking method based on Camshift and SIFT was proposed in this paper. In this method, SIFT feature points, which were used to construct the color histogram and the color probability distribution, were extracted from the target region first. Then SIFT points were also extracted from the search region and these two sets of SIFT points were matched. Since the proposed method used the matched SIFT points to properly guide the location of targets, experimental results show that with the new method some more accurate and robust tracking results have been obtained.

Sign in / Sign up

Export Citation Format

Share Document