Object Tracking Using Adaptive Diffusion Flow Active Model

Object tracking is one of the most important topics in the fields of image processing and computer vision. Object tracking is the process of finding interesting moving objects and following them from frame to frame. In this research, Active models–based object tracking algorithm is introduced. Active models are curves placed in an image domain and can evolve to segment the object of interest. Adaptive Diffusion Flow Active Model (ADFAM) is one the most famous types of Active Models. It overcomes the drawbacks of all previous versions of the Active Models specially the leakage problem, noise sensitivity, and long narrow hols or concavities. The ADFAM is well known for its very good capabilities in the segmentation process. In this research, it is adopted for segmentation and tracking purposes. The proposed object tracking algorithm is initiated by detecting the target moving object manually. Then, the ADFAM convergence of the current video frame is reused as an initial estimation for the next video frame and so on. The proposed algorithm is applied to several video sequences, different in terms of the nature of the object, the nature of the background, the speed of the object, object motion direction, and the inter-frame displacement. Experimental results show that the proposed algorithm performed very well and successfully tracked the target object in all different cases.

Download Full-text

The Application of Image Processing to Solve Occlusion Issue in Object Tracking

MATEC Web of Conferences ◽

10.1051/matecconf/201815203001 ◽

2018 ◽

Vol 152 ◽

pp. 03001

Author(s):

Yun Zhe Cheong ◽

Wei Jen Chew

Keyword(s):

Object Tracking ◽

Moving Objects ◽

Object Segmentation ◽

Target Object ◽

Occlusion Handling ◽

Multiple Objects ◽

The Road ◽

Occluded Objects ◽

On The Road ◽

Tracking Object

Object tracking is a computer vision field that involves identifying and tracking either a single or multiple objects in an environment. This is extremely useful to help observe the movements of the target object like people in the street or cars on the road. However, a common issue with tracking an object in an environment with many moving objects is occlusion. Occlusion can cause the system to lose track of the object being tracked or after overlapping, the wrong object will be tracked instead. In this paper, a system that is able to correctly track occluded objects is proposed. This system includes algorithms such as foreground object segmentation, colour tracking, object specification and occlusion handling. An input video is input to the system and every single frame of the video is analysed. The foreground objects are segmented with object segmentation algorithm and tracked with the colour tracking algorithm. An ID is assigned to each tracked object. Results obtained shows that the proposed system is able to continuously track an object and maintain the correct identity even after is has been occluded by another object.

Download Full-text

Information Technology in Multi-Object Tracking Based on Bilateral Structure Tensor Corner Detection for Mutual Occlusion

Advanced Materials Research ◽

10.4028/www.scientific.net/amr.977.502 ◽

2014 ◽

Vol 977 ◽

pp. 502-506

Author(s):

Ying Hong Xie ◽

Xiao Wei Han ◽

You Guo He

Keyword(s):

Information Technology ◽

Object Tracking ◽

Nearest Neighbor ◽

Tracking Algorithm ◽

Structure Tensor ◽

Corner Detection ◽

K Nearest Neighbor ◽

Motion Direction ◽

Tracking Process ◽

K Nearest Neighbor Algorithm

For mutual occlusion problem in multi-object tracking process, a novel tracking algorithm based on bilateral structure tensor corner detection is proposed, which can separate the objects correctly when they experience mutual occlusion. Firstly, it gains the information of each object corners. Secondly, when occlusion occurs, it makes use of K nearest neighbor algorithm combining with the nearest algorithm to classify the corners in occlusion region. Finally, the multi-object tracking algorithm is proposed. The experimental results show that the proposed method can separate the objects correctly and track the objects effectively, when they experience mutual occlusion, even the object changes its motion direction after occlusion.

Download Full-text

Object Tracking Based on Dual-View Stereo System

Advanced Materials Research ◽

10.4028/www.scientific.net/amr.850-851.780 ◽

2013 ◽

Vol 850-851 ◽

pp. 780-783

Author(s):

Jian De Fan ◽

Jiang Bo Zhu

Keyword(s):

Computer Vision ◽

Optical Flow ◽

Object Tracking ◽

Moving Objects ◽

Current Method ◽

Research Area ◽

Tracking Algorithm ◽

Video Sequences ◽

Feature Points ◽

Motion Trajectories

Tracking moving objects in dual-view stereo system is becoming a hot research area in computer vision. To capture the moving objects pixels more accurately, we proposed a new object tracking algorithm which first compute moving objects feature points and then match these points, finally connect the matching feature points and get objects motion trajectories. The algorithm was tested in the video sequences with resolution 640×480 and 768×576 individually. The results show that the algorithm is more robust and the trajectories of the moving objects tracked with our method are more accurate compared with current method of L-K optical flow.

Download Full-text

Object Tracking Algorithm Based on Block LAB Feature Histogram and Particle Filter

Advanced Materials Research ◽

10.4028/www.scientific.net/amr.485.193 ◽

2012 ◽

Vol 485 ◽

pp. 193-199

Author(s):

Ming Sun ◽

Jia Wei Li

Keyword(s):

Particle Filter ◽

Object Tracking ◽

Real Time ◽

Moving Objects ◽

Tracking Algorithm ◽

Color Histogram ◽

Complex Environments ◽

Excellent Performance ◽

Real Time Tracking ◽

Tracking Objects

In order to improve real-time object tracking effect when tracking objects are partly covered or mixed by different backgrounds, and even under the conditions of changed illuminations, in this paper, we proposed an object tracking algorithm based on block LAB feature histogram and particle filter. To demonstrate new algorithm’s excellent performance, we carried several compared experiments when objects moved under different conditions such as changed illuminations, mixed backgrounds and so forth. Experiment results show that tracking objects are often lost by using tracking algorithm based on traditional features such as color histogram, but moving objects under various and complex environments could be correctly tracked by using real-time tracking algorithm proposed in this paper.

Download Full-text

How We See and Recognize Object Motion

Conscious Mind, Resonant Brain ◽

10.1093/oso/9780190070557.003.0008 ◽

2021 ◽

pp. 289-336

Author(s):

Stephen Grossberg

Keyword(s):

Decision Making ◽

Long Range ◽

Motion Perception ◽

Apparent Motion ◽

Feature Tracking ◽

Moving Objects ◽

Object Motion ◽

Motion Direction ◽

Neurophysiological Data ◽

Object Depth

This chapter explains why visual motion perception is not just perception of the changing positions of moving objects. Computationally complementary processes process static objects with different orientations, and moving objects with different motion directions, via parallel cortical form and motion streams through V2 and MT. The motion stream pools multiple oriented object contours to estimate object motion direction. Such pooling coarsens estimates of object depth, which require precise matches of oriented stimuli from both eyes. Negative aftereffects of form and motion stimuli illustrate these complementary properties. Feature tracking signals begin to overcome directional ambiguities due to the aperture problem. Motion capture by short-range and long-range directional filters, together with competitive interactions, process feature tracking and ambiguous motion directional signals to generate a coherent representation of object motion direction and speed. Many properties of motion perception are explained, notably barberpole illusion and properties of long-range apparent motion, including how apparent motion speed varies with flash interstimulus interval, distance, and luminance; apparent motion of illusory contours; phi and beta motion; split motion; gamma motion; Ternus motion; Korte’s Laws; line motion illusion; induced motion; motion transparency; chopsticks illusion; Johannson motion; and Duncker motion. Gaussian waves of apparent motion clarify how tracking occurs, and explain spatial attention shifts through time. This motion processor helps to quantitatively simulate neurophysiological data about motion-based decision-making in monkeys when it inputs to a model of how the lateral intraparietal, or LIP, area chooses a movement direction from the motion direction estimate. Bayesian decision-making models cannot explain these data.

Download Full-text

Analysis of visual object tracking algorithms for real-time systems

Advanced Information Technology ◽

10.17721/ait.2021.1.08 ◽

2021 ◽

pp. 59-65

Author(s):

Mykola Moroz ◽

Denys Berestov ◽

Oleg Kurchenko

Keyword(s):

Object Tracking ◽

Real Time ◽

Video Stream ◽

Target Object ◽

Tracking Algorithm ◽

Visual Object ◽

Security Systems ◽

Visual Support ◽

Tracking Object ◽

Time Systems

The article analyzes the latest achievements and decisions in the process of visual support of the target object in the field of computer vision, considers approaches to the choice of algorithm for visual support of objects on video sequences, highlights the main visual features that can be based on tracking object. The criteria that influence the choice of the target object-tracking algorithm in real time are defined. However, for real-time tracking with limited computing resources, the choice of the appropriate algorithm is crucial. The choice of visual tracking algorithm is also influenced by the requirements and limitations for the monitored objects and prior knowledge or assumptions about them. As a result of the analysis, the Staple tracking algorithm was preferred, according to the criterion of speed, which is a crucial indicator in the design and development of software and hardware for automated visual support of the object in real-time video stream for various surveillance and security systems, monitoring traffic, activity recognition and other embedded systems.

Download Full-text

Quantized Kalman Filter-Based Pattern Matching for Detection and Tracking of Moving Objects

International Journal of Engineering and Advanced Technology - Regular Issue ◽

10.35940/ijeat.a9839.109119 ◽

2019 ◽

Vol 9 (1) ◽

pp. 3842-3851

Keyword(s):

Kalman Filter ◽

Object Tracking ◽

Pattern Matching ◽

Kalman Filtering ◽

Moving Objects ◽

Moving Object ◽

Video Frame ◽

Top Down ◽

Matching Accuracy ◽

Detection And Tracking

Detection And Tracking Of Multiple Moving Objects From A Sequence Of Video Frame And Obtaining Visual Records Of Objects Play An Important Role In The Video Surveillance Systems. Transform And Filtering Technique Designed For Video Pattern Matching And Moving Object Detection, Failed To Handle Large Number Of Objects In Video Frame And Further Needs To Be Optimized. Several Existing Methods Perform Detection And Tracking Of Moving Objects. However, The Performance Efficiency Of The Existing Methods Needs To Be Optimized To Achieve More Robust And Reliable Detection And Tracking Of Moving Objects. In Order To Improve The Pattern Matching Accuracy, A Quantized Kalman Filter-Based Pattern Matching (Qkf-Pm) Technique Is Proposed For Detecting And Tracking Of Moving Objects. The Present Phase Includes Three Functionalities: Top-Down Approach, Kernel Pattern Segment Function And Kalman Filtering. First, The Top-Down Approach Based On Kalman Filtering (Kf) Technique Is Performed To Detect The Chromatic Shadows Of Objects. Next, Kernel Pattern Segment Function Creates The Seed Points For Detecting Moving Object Pattern. Finally, Object Tracking Is Performed Using The Proposed Quantized Kalman Filter Based On The Center Of Seed Point Affinity Feature Values Are Used To Track The Moving Objects In A Particular Region Using The Minimum Bounding Box Approach. Experimental Results Reveals That The Proposed Qkf-Pm Technique Achieves Better Performance In Terms Of True Detection Rate, Pattern Matching Accuracy, Pattern Matching Time, And Object Tracking Accuracy With Respect To The Number Of Video Frames Per Second.

Download Full-text

PFCHA: A New Moving Object Tracking Algorithms Based on Particle Filter and Histogram

Applied Mechanics and Materials ◽

10.4028/www.scientific.net/amm.110-116.3343 ◽

2011 ◽

Vol 110-116 ◽

pp. 3343-3350

Author(s):

Qi Yang ◽

Jia Fu Jiang

Keyword(s):

Particle Filter ◽

Object Tracking ◽

Target Tracking ◽

Moving Objects ◽

Tracking Algorithm ◽

Moving Target ◽

Trace Detection ◽

Complex Background ◽

Motion Characteristics ◽

The Impact

The complexity of the video background of moving target tracking algorithm led to the robustness of the important reasons is not high for the limitations of existing algorithms, a framework based on the movement of particle filter tracking algorithm. In order to reduce the impact of occlusion for the algorithm, the algorithm of moving objects make full use of color and motion characteristics of moving target detection, and to avoid the interference of the complex background, within the framework of particle filter in the object color histogram analysis. Finally, given an effective comparison of the calculation. Experimental results show that particle filter based target tracking algorithm can effectively remove the interference of the complex background, the context for any trace detection of high robustness.

Download Full-text

A Two-Stage Data Association Approach for 3D Multi-Object Tracking

Sensors ◽

10.3390/s21092894 ◽

2021 ◽

Vol 21 (9) ◽

pp. 2894

Author(s):

Minh-Quan Dao ◽

Vincent Frémont

Keyword(s):

Object Detection ◽

Object Tracking ◽

Moving Objects ◽

Data Association ◽

Autonomous Driving ◽

Tracking Accuracy ◽

Two Stage ◽

Bipartite Matching ◽

3D Object ◽

3D Object Detection

Multi-Object Tracking (MOT) is an integral part of any autonomous driving pipelines because it produces trajectories of other moving objects in the scene and predicts their future motion. Thanks to the recent advances in 3D object detection enabled by deep learning, track-by-detection has become the dominant paradigm in 3D MOT. In this paradigm, a MOT system is essentially made of an object detector and a data association algorithm which establishes track-to-detection correspondence. While 3D object detection has been actively researched, association algorithms for 3D MOT has settled at bipartite matching formulated as a Linear Assignment Problem (LAP) and solved by the Hungarian algorithm. In this paper, we adapt a two-stage data association method which was successfully applied to image-based tracking to the 3D setting, thus providing an alternative for data association for 3D MOT. Our method outperforms the baseline using one-stage bipartite matching for data association by achieving 0.587 Average Multi-Object Tracking Accuracy (AMOTA) in NuScenes validation set and 0.365 AMOTA (at level 2) in Waymo test set.

Download Full-text