target tracking system

download target tracking system

of 5

Transcript of target tracking system

  • 8/10/2019 target tracking system

    1/5

    INGHUA SCIENCE AND TECHNOLOGY

    SN 1007-0214 55/67 pp343-347

    lume13,NumberS1,October2008

    A Target Tracking System for Applications in Hydraulic Engineering

    SHEN Qiaonan (), AN Xuehui ()**

    Department of Hydraulic and Hydropower Engineering, Tsinghua University, Beijing 100084, China

    Abstract:A new type of digital video monitoring system (DVMS) named user defined target tracking system

    (UDTTS), was developed based on the digital image processing (DIP) technology and the practice demands

    of construction site management in hydraulic engineering. The position, speed, and track of moving targets

    such as humans and vehicles, which could be calculated by their locations at anytime in images basically,

    were required for management. The proposed algorithm, dependent on the context-sensitive moving infor-

    mation of image sequences which was much more than one or two images provided, compared the blobs

    properties in current frame to the trajectories of targets in the previous frames and then corresponded them.

    The processing frame rate is about 10fps with the image 240-by-120 pixels. Experimental results show that

    position, direction, and speed measurements have an accuracy level compatible with the manual work. The

    user-define process makes the UDTTS available to the public whenever appropriate.

    Key words: target tracking system; digital image processing; user-defined; consecutive trajectory

    ntroduction

    is widely recognized that hydraulic construction en-

    neering is information intensive and complex indus-

    y. Present trends in the hydraulic construction engi-

    ering have heightened the need for effective and ef-

    ient collecting, monitoring and analysis the con-

    uction progress data. In recent years, the use of digi-

    video monitoring system (DVMS) in the surveil-nce phase of a project is rapidly growing which im-

    oves the progress controlling, safety monitoring and

    ork coordination during entire project[1]

    .

    However, information within thousands of digital

    deos and images stored for a project from the DVMS

    uld not be obtained automatically.

    A large number of components and their features

    ed to be inspected on construction sites[2-3]

    . Many of

    ese features need to be assessed based on tight toler-ces, requiring that inspections be extremely accurate.

    At the same time, inspection resources, such as the

    time that inspectors can spend on site, are limited.

    Therefore, inspectors can benefit from emerging tech-

    nologies that improve the efficiency of data collection

    while on site, and from visualization technologies that

    improve the effectiveness and efficiency of inspection

    tasks using this data.

    The capability to automatically identify objects from

    images through many methodologies is a product of

    the technological breakthroughs in the area of digital

    image processing (DIP)[4,5]

    .

    Detection and tracking of targets in construction site

    is not only a single object tracking problem, but also a

    multi-object tracking problem. Numerous approaches[6]

    for multi-object tracking have been proposed. But it is

    still a very different and more challenging problem. In

    addition to the normal frame-to-frame following of a

    salient area, the system must be able to handle occur-

    rences, disappearances, crossing and other complicated

    events related to multiple moving targets. Features[7-12]

    such as color, texture, shape, and motion properties are

    used for tracking.

    In this study, a new type of DVMS named user

    Received: 2008-05-30

    To whom correspondence should be addressed.

    E-mail: [email protected]; Tel: 86-10-62794285

  • 8/10/2019 target tracking system

    2/5

    Tsinghua Science and Technology,October 2008, 13(S1): 343-347344

    fined target tracking system (UDTTS) was proposed

    d developed based on the DIP technology and the

    actice demands of construction site management in

    draulic engineering. And a new algorithm was pro-

    sed for multi-object tracking, dependent on blob

    operties and context-sensitive motion information.

    System Overview

    he system called UDTTS includes four parts: User-

    fined part, data preprocessing, moving object detec-

    n and tracking. The input data is a video file or the

    eam of images captured by a stationary digital video

    ounted on a horizontal gantry or on a tripod and in

    atic positions at construction site.

    1

    User-defined process

    his system can do many aspects of management by

    er-define process. Users can define the application,

    ch as vehicle flow, human flow, grinding variables

    three steps. Images including targets and static

    ckground should be provided to the UDTTS. Firstly,

    nerate the initial background model when the back-

    ound image is input; secondly, define a target on the

    get image captured on construction site; thirdly, de-ne the controlling conditions that the target must sat-

    y; finally, define an output format. So the definition

    an application is finished.

    2

    Application analysis

    oving targets such as vehicles, humans, and other

    ngs at construction site have variable colors, sizes,

    apes, speeds, and directions. Their features can be

    lized to detect and track them. As is shown in Fig. 1,application can be worked out from a targets trajec-

    ry which consists of its positions at sequential time.

    he problem is how to know the positions of a target at

    y time from the streams of color image. In the

    Fig. 1 Application analysis

    UDTTS, after the user-define process, the video cap-

    tured on construction site is input to be processed. The

    procedure performs several images processing tasks to

    detect and tracking moving objects in the scene. The

    result can be output as user-define format.

    2 Tracking Method

    The purpose of the tracking part is to detect moving

    objects from the video stream and collect appropriate

    data of their routes. Tracking is usually performed in

    the context of higher-level applications that require the

    location and/or shape of the object in every frame.

    Typically, assumptions are made to constrain the track-

    ing problem in the context of a particular application.

    In its simplest form, tracking can be defined as the

    problem of estimating the trajectory of an object in the

    image plane as it moves around a scene.

    The task of detecting and tracking moving objects

    from video deals with the problem of extracting mov-

    ing objects (foreground-background separation) and

    generating corresponding persistent trajectories. In the

    case of multiple objects in the scene, the tracking task

    is equivalent with the task of solving the correspon-

    dence problem. At each frame a set of trajectories and

    a set of measured objects (blobs) are available. Each

    object is identified by finding the matching trajectory.

    2.1

    Detection of moving objects

    Detection of moving objects in video streams is the

    first relevant step of information extraction in many

    computer vision applications. Aside from the intrinsic

    usefulness of being able to segment video streams into

    moving and background components, detecting mov-ing objects provides a focus of attention for recogni-

    tion, classification, and activity analysis, making these

    later steps more efficient.

    At hardware level, color images are usually captured,

    stored and displayed using elementary R, G, B compo-

    nent images. The color images read from the frame

    grabber are transformed to gray scale images with only

    luminance information preserved in order to reduce the

    computational load and to guarantee adequate framerate (around 10 fps) for tracking. Each incoming frame

    goes through four successive image processing stages

    where the raw intensity data is reduced to a compact

    set of features which can be used for the matching

  • 8/10/2019 target tracking system

    3/5

    SHEN Qiaonan () et alA Target Tracking System for Applications in Hydraulic Engineering 345

    ethod. These four stages are gray-scale transforma-

    n, background subtraction, threshold segmentation

    d connected component labeling as is shown in

    g. 2.

    Fig. 2 The digital image processing steps

    Motion detection is started by computing a pixel

    sed absolute difference between each incoming

    ame and the static background frame provided by us-

    s. The pixels are assumed to contain motion if the

    solution difference exceeds a predefined threshold

    vel. As a result, a binary image is formed where ac-

    e pixels are labeled with 1 and non-active ones

    th 0.

    The figures directly extracted from the resulting bi-

    ry image are typically fragmented to multiple seg-

    ents. In order to avoid this, a morphological closing

    eration with a 3-by-3 square kernel is applied to the

    age. As a result, small gaps between the isolated

    gments are erased and the regions are merged.

    After closing, we use a connected component analy-[13]

    followed by region area detection in this stage.

    he regions with a smaller area than the predefined

    reshold are now discarded.

    Position and area of each blob are detected in local

    odel of individual frame. After detection, the objects

    a local model of single frame must be integrated to

    e trajectories in a world model of all frames through

    atching method.

    2

    Tracking of moving objects

    acking is needed for determining the object corre-

    ondence between frames. In our approach, the main

    cked feature is the object trajectory which is con-

    cutive in frame sequences. Since the speed of the

    objects at construction site is not too fast, we assume

    that the blob in current frame and its corresponding tra-

    jectory in the previous frames overlap. The object cen-

    troid and dynamic information are used for tracking.

    The speed and direction of the object generated by the

    previous trajectory are stored in the world model of all

    frames. They are also useful features for matching.In general, high occurrences of objects that visually

    overlap cause difficulties for a tracking system. Since

    blob generation of moving objects is based on con-

    nected component analysis, touching objects generate a

    single merged object, where pixel classification, i.e., to

    which original blob individual pixels belong, is hard to

    resolve. This lead to the problem that in a merged state

    individual tracks cannot be updated. To overcome this

    problem, we propose a solution using a technique,which generates plausible trajectories of the objects in

    a merged state by performing matching between ob-

    jects entering and leaving the merged state. The match-

    ing is based on the kinematic smoothness constraint.

    The method is presented in section 2.3.

    In the first frame, each blob generates a trajectory

    with the following attributes: area, speed, direction and

    status. Consecutive judgement is used for matching,

    which is described in section 2.3.The scheme of the tracking algorithm is outlined as

    follows.

    Step 1 If a blob is exactly matched to one existing

    trajectory, the trajectory properties (area, speed, direc-

    tion, and status) are updated.

    Step 2 If a blob matches two trajectories, crossing

    happens. Set the status of these trajectories crossing.

    Then do not process them until splitting happens.

    Step 3

    If a trajectory matches two blobs, splittinghappens. Find the partner trajectory and compare them

    to these two blobs. Update the two trajectories proper-

    ties.

    Step 4

    If a none-matched blob is found, a new tra-

    jectory is generated.

    Step 5 In case of detecting a non-matched trajec-

    tory, exiting or failure of the blob detection happens. If

    the trajectory tends to be out of the view, maybe exit-

    ing is right; or leave it to be processed in next frame.

    2.3

    Consecutive judgement

    Consecutive judgement: As is shown in Fig. 3, if a

    blob with solid line in current frame and a trajectory

  • 8/10/2019 target tracking system

    4/5

    Tsinghua Science and Technology,October 2008, 13(S1): 343-347346

    th dotted line overlap, we say they are consecutive,

    herwise, they are inconsecutive.

    Fig. 3

    Consecutive and inconsecutive trajectory

    In the case of inconsecutive trajectory, these features

    maximum distance, limited speed, and correlative di-

    ction) are used for matching (conditions shown in

    g. 4).

    Fig. 4

    Inconsecutive trajectory conditions

    If a trajectory is only generated by one blob, speed

    d direction are not effective values. The distance dij

    tween current blob centroid and the previous blob

    ntroid should fulfill the condition as

    dijd X (1)

    here Xd denotes maximum distance an object can

    ove in a certain interval, i, jare the frame number.

    If a trajectory is generated by more than two blobs,

    eed and direction can be used for matching. If the

    rrent speed vand the direction correlation described

    are in the acceptable range, i.e.,1 1

    VX VX

    1 1

    VY VY

    (1 )* (1 )*

    (1 )* (1 )*

    n n n

    x x x

    n n n

    y y y

    x V V x V

    x V V x V

    (2)

    here Vx is the speed in X-axis, Vy is the speed in Y-

    is, nis the frame number,xVXandxVYare predefined

    ios in (0,1);

    1cos cos 1 (3)

    here is the angle between the current direction and

    e previous, 1 is the predefined angel in (90, 90),

    e blob and the trajectory match each other, otherwise

    ey do not.

    As described above, when blobs overlap the obser-

    vation of a single merged blob does not allow recon-

    structing the trajectories of the original entering blobs.

    Just add the blob to these trajectories for the latter con-

    secutive judgement. Remember the frame number i

    and the time at which crossing happens. When splitting

    happens at frame k, direction consistence and correla-tive speed are used for matching the blobs and the tra-

    jectories based on the kinematic smoothness constraint.

    In the case of entering or exiting, the blob must be

    near to the boundary of the processing area.

    3 An Example

    The tracking system UDTTS has been applied to two

    video files captured from Xiangjiaba dam to track ve-

    hicles. One of the test sequences contains one object

    and the other one contains multiple objects occurring

    entering, exiting and crossing events. The static back-

    ground is provided to define the processing area (the

    rectangle in Fig. 5), and the targets area is obtained

    before processing. Main parameters of algorithm im-

    plementation: Windows XP, VC++ 6.0, CPU AMD

    Athlon 2.01 GHz, and memory 1.00 GB. The process-

    ing frame rate is about 8 fps, while the image size is

    240 by 120. The accuracy and stability of the system

    depend on these parameters which are predefined

    initially.

    The second sequence contains 4 vehicles generating

    1 entering, 4 exiting and 3 crossing events. The track-

    ing results such as the centroid sequence and the trajec-

    tory of each vehicle are shown in Fig. 5. 4 frames at

    1st second, 3rd second, 7-th second and 26-th second

    are listed on the left. Crossing event between vehicle 2

    and 3 occurs at T2, vehicle 1 and 3 cross at T3, and ve-

    hicle1 and vehicle 4. Vehicle 3 is moving out of the

    processing area at T4. Vehicle 1 disappears at T4, and

    vehicle 2 has moved out of the processing area from T3.

    Vehicle 4 appears after T3 and leaves the processing

    area finally.

    A qualitative summary of the observed events is

    summarized in Table 1.

    Table 1

    Critical events processed by tracking method

    ItemsEntering

    events

    Exiting

    event

    Crossing

    events

    Test results 1 4 3

    Actual situations 1 4 3

  • 8/10/2019 target tracking system

    5/5

    SHEN Qiaonan () et alA Target Tracking System for Applications in Hydraulic Engineering 347

    Fig. 5

    Tracking results for the video containing multiple moving objects

    Conclusions and Future Work

    e have presented UDTTS for real-time moving target

    cking. Real-time multi-object detection and tracking

    gorithm were developed using consecutive trajectory

    d correlative motion information. Experimental re-lts show that position, direction, and speed meas-

    ements have an accuracy level compatible with the

    anual work.

    Adaptive background model will be developed and

    e efficiency of the algorithm will be improved. More

    mplex scenes should be tested for the UDTTS. Al-

    ough the UDTTS is developed to meet the needs of

    nstruction site management, it is available to the

    blic whenever appropriate.

    eferences

    Memon Z A, Abd Majid M Z, Mustaffar M. An automatic

    project progress monitoring model by integrating Auto-

    CAD and digital photos. In: Proceeding of Computing in

    Civil Engineering (CCE) ASCE. Cancun, Mexico, 2005.

    Soibelman L, Brilakis I. Identification of material from

    construction site images using content based image re-

    trieval techniques. In: Proceedings of Computing in Civil

    Engineering ASCE. Cancun, Mexico, 2005.

    Teizer J, Caldas C H, Haas C. Real-time three-dimensional

    occupancy grid Modeling for the detection and tracking of

    construction resources.ASCE Journal of Construction En-

    gineering and Management, 2007, 133(11): 880-888.

    Yilmaz A, Javed O, Shah M. Object tracking: A survey.

    ACM Journal of Computing Surveys, 2006, 38(4): 1-45.

    [5] Beleznai C, Schlgl T, Wachmann B, et al. Tracking multi-

    ple objects in complex scenes. In: Leberl F, Fraundorfer F,

    eds, Vision with Non-Traditional Sensors. Proceeding of of

    26th Workshop of the Austrian Association for Pattern

    Recognition. Austrian Computer Society, 2002, 160:

    175-182.[6] Yilmaz A, Javed O, Shah M. Object tracking: A survey.

    ACM Comput. Surv. 2006, 4(38): 1-45.

    [7] Takala V, Pietikainen M. Multi-object tracking using color,

    texture and motion. In: IEEE Conference on Computer Vi-

    sion and Pattern Recognition. 2007: 1-7.

    [8] Veeraraghavan H, Schrater P, Papanikolopoulos N. Robust

    target detection and tracking through integration of motion,

    color, and geometry. Computer Vision and Image Under-

    standing, 2006, 103(2): 121-138.

    [9] Vermaak J, Godsill S J, Perez P. Monte Carlo filtering for

    multi-target tracking and data association. IEEE Transac-

    tions on Aerospace and Electronic Systems, 2005, 41(1):

    309-332.

    [10]Wang Shuan, Ai Haizhou, He Kezhong. Difference-image-

    based multiple motion targets detection and tracking.Jour-

    nal of Image and Graphic, 1999, 4(6): 470-475.

    (in Chinese)

    [11]Wan Qin, Wang Yaonan. Research and implementation of

    detecting and tracking multiple moving objects method.

    Application Research of Computers, 2007, 1: 199-202.

    (in Chinese)

    [12]Ge Jiaqi, Li Bo, Chen Qimei. A region-based vehicle track-

    ing algorithm under occlusion.Journal of NanJing Univer-

    sity (Natural Sciences), 2007, 43(1): 66-72. (in Chinese)

    [13]Gao Hongbo, Wang Weixing. New connected component

    labeling algorithm for binary image. Computer Applica-

    tions, China, 2007, 27(11): 2776-2777. (in Chinese)