22
Introduction Concept and implementation Evaluation Conclusions and future work Collaborative, Context Based Activity Control Method for Camera Networks Marek Kraft 1 Michal Fularz 1 Adam Schmidt 1 1 Poznań University of Technology Institute of Control and Information Engineering November 3, 2015 M. Kraft, M. Fularz, A. Schmidt Collaborative, Context Based Activity Control Method...

Collaborative, Context Based Activity Control Method for Camera Networks

Embed Size (px)

Citation preview

Page 1: Collaborative, Context Based Activity Control Method for Camera Networks

IntroductionConcept and implementation

EvaluationConclusions and future work

Collaborative, Context Based Activity ControlMethod for Camera Networks

Marek Kraft1 Michał Fularz1 Adam Schmidt1

1Poznań University of TechnologyInstitute of Control and Information Engineering

November 3, 2015

M. Kraft, M. Fularz, A. Schmidt Collaborative, Context Based Activity Control Method...

Page 2: Collaborative, Context Based Activity Control Method for Camera Networks

IntroductionConcept and implementation

EvaluationConclusions and future work

Table of contentsGoal and motivation

1 IntroductionTable of contentsGoal and motivation

2 Concept and implementationKey conceptsBackground subtractionNode activationCommunication and coordination

3 EvaluationHardware platformTest setupResults

4 Conclusions and future work

M. Kraft, M. Fularz, A. Schmidt Collaborative, Context Based Activity Control Method...

Page 3: Collaborative, Context Based Activity Control Method for Camera Networks

IntroductionConcept and implementation

EvaluationConclusions and future work

Table of contentsGoal and motivation

GoalDevelop a conceptually simple yet effective, collaborative methodfor constraining the rate of communication and power consumptionacross the whole camera network.

MotivationAs the average number of cameras in a network increases:

automated processing becomes a necessity,continuous image streaming and data transmission is a seriousburden to the communication infrastructure,combined power consumption of the camera networks goes up,most previous work is focused on the design of network nodesor single node management.

M. Kraft, M. Fularz, A. Schmidt Collaborative, Context Based Activity Control Method...

Page 4: Collaborative, Context Based Activity Control Method for Camera Networks

IntroductionConcept and implementation

EvaluationConclusions and future work

Table of contentsGoal and motivation

M. Kraft, M. Fularz, A. Schmidt Collaborative, Context Based Activity Control Method...

Page 5: Collaborative, Context Based Activity Control Method for Camera Networks

IntroductionConcept and implementation

EvaluationConclusions and future work

Key conceptsBackground subtractionNode activationCommunication and coordination

Key concepts:compute the activity level for each individual sensor locally,based on scene motion level,scene motion level is calculated based on backgroundsubtraction,the activation level of each sensor node depends on the localactivity, but also on the activity of its neighbors,decision on the kind of computation performed by the sensornode and its other activities are made depending on theactivation level.

M. Kraft, M. Fularz, A. Schmidt Collaborative, Context Based Activity Control Method...

Page 6: Collaborative, Context Based Activity Control Method for Camera Networks

IntroductionConcept and implementation

EvaluationConclusions and future work

Key conceptsBackground subtractionNode activationCommunication and coordination

Background subtractionForeground is the absolute value of the difference between thecurrent frame and the background model.Additional post-processing applied to foreground image.

N. J. McFarlane et. al., Segmentation and tracking of piglets inimages. Machine vision and applications, 8(3), pp. 187-193, 1995

S. Brutzer et. al., Evaluation of background subtraction techniquesfor video surveillance, IEEE Conf. on Computer Vision and PatternRecognition (CVPR) 2011, pp. 1937-1944

M. Kraft, M. Fularz, A. Schmidt Collaborative, Context Based Activity Control Method...

Page 7: Collaborative, Context Based Activity Control Method for Camera Networks

IntroductionConcept and implementation

EvaluationConclusions and future work

Key conceptsBackground subtractionNode activationCommunication and coordination

Background subtractionBackground model:

if the intensity value Ix ,y of currently investigated pixel of thecurrent frame I is greater than the value of the correspondingbackground model pixel Bx ,y , the value of Bx ,y is increased,if the intensity value Ix ,y of currently investigated pixel of thecurrent frame I is smaller than the value of the correspondingbackground model pixel Bx ,y , the value of Bx ,y is decreased,if the intensity value Ix ,y of currently investigated pixel of thecurrent frame I has the same value as the correspondingbackground model pixel Bx ,y , the value of Bx ,y remainsunchanged.

M. Kraft, M. Fularz, A. Schmidt Collaborative, Context Based Activity Control Method...

Page 8: Collaborative, Context Based Activity Control Method for Camera Networks

IntroductionConcept and implementation

EvaluationConclusions and future work

Key conceptsBackground subtractionNode activationCommunication and coordination

Background subtraction

M. Kraft, M. Fularz, A. Schmidt Collaborative, Context Based Activity Control Method...

Page 9: Collaborative, Context Based Activity Control Method for Camera Networks

IntroductionConcept and implementation

EvaluationConclusions and future work

Key conceptsBackground subtractionNode activationCommunication and coordination

Background subtraction

M. Kraft, M. Fularz, A. Schmidt Collaborative, Context Based Activity Control Method...

Page 10: Collaborative, Context Based Activity Control Method for Camera Networks

IntroductionConcept and implementation

EvaluationConclusions and future work

Key conceptsBackground subtractionNode activationCommunication and coordination

Block schematic of a single nodeThe percentages of foreground pixels in the local sensor nodeand its defined network neighbors are used as inputs.2nd order inertia applied for low-pass filtering.

M. Kraft, M. Fularz, A. Schmidt Collaborative, Context Based Activity Control Method...

Page 11: Collaborative, Context Based Activity Control Method for Camera Networks

IntroductionConcept and implementation

EvaluationConclusions and future work

Key conceptsBackground subtractionNode activationCommunication and coordination

Connections in the networkA central coordinator is responsible for network-wide activitycontrol which based on individual sensor states.The sensors provide the coordinator with the information ontheir individual state and transmit images if requested.

M. Kraft, M. Fularz, A. Schmidt Collaborative, Context Based Activity Control Method...

Page 12: Collaborative, Context Based Activity Control Method for Camera Networks

IntroductionConcept and implementation

EvaluationConclusions and future work

Key conceptsBackground subtractionNode activationCommunication and coordination

Connections in the networkThe neighborhood information is handled by the coordinator.The coordinator stores the information on the mutual relationsof network nodes (in the form of gain values), forming virtualinter-sensor connections.

M. Kraft, M. Fularz, A. Schmidt Collaborative, Context Based Activity Control Method...

Page 13: Collaborative, Context Based Activity Control Method for Camera Networks

IntroductionConcept and implementation

EvaluationConclusions and future work

Key conceptsBackground subtractionNode activationCommunication and coordination

Connections in the networkBased on the node activity levels, the coordinator computesthe additional portion of the input value for each node.This enables the use of network-wide information forcollaborative node activity control.

M. Kraft, M. Fularz, A. Schmidt Collaborative, Context Based Activity Control Method...

Page 14: Collaborative, Context Based Activity Control Method for Camera Networks

IntroductionConcept and implementation

EvaluationConclusions and future work

Hardware platformTest setupResults

Hardware platform – the PiCam

off-the-shelf 1st generationRaspberry Pi withARM1176JZF-S CPU and 512MB RAMstandard Raspberry Pi cameraUSB WiFi cardpowerbank power supply forportabilityfisheye lens for enhanced field ofviewruns Arch Linux

M. Kraft, M. Fularz, A. Schmidt Collaborative, Context Based Activity Control Method...

Page 15: Collaborative, Context Based Activity Control Method for Camera Networks

IntroductionConcept and implementation

EvaluationConclusions and future work

Hardware platformTest setupResults

PiCam operating modesEach node was configured for two operating modes.The node switches to performance mode if activation is aboveTA, or switches back to powersave mode otherwise.The sampling (and processing) frequency is 10 [Hz] inperformance and 1 [Hz] in powersave mode.

parameter performance mode powersave mode

CPU clock [MHz] 1,000 300

RAM clock [MHz] 500 150

GPU clock [MHz] 500 150

Current draw [A] 0.5 0.4

M. Kraft, M. Fularz, A. Schmidt Collaborative, Context Based Activity Control Method...

Page 16: Collaborative, Context Based Activity Control Method for Camera Networks

IntroductionConcept and implementation

EvaluationConclusions and future work

Hardware platformTest setupResults

The test setup – five cameras placed in typical office space:

M. Kraft, M. Fularz, A. Schmidt Collaborative, Context Based Activity Control Method...

Page 17: Collaborative, Context Based Activity Control Method for Camera Networks

IntroductionConcept and implementation

EvaluationConclusions and future work

Hardware platformTest setupResults

Test scenarioA person moves from room 2 through the corridor to room 1 andback two times. The scenario lasts approximately 5 minutes.

Gain table

camera PiCam01 PiCam02 PiCam03 PiCam04 PiCam05

PiCam01 0.0 0.5 0.2 0.0 0.0PiCam02 0.5 0.0 0.5 0.1 0.0PiCam03 0.1 0.5 0.0 0.5 0.2PiCam04 0.0 0.1 0.5 0.0 0.5PiCam05 0.0 0.0 0.2 0.5 0.0

M. Kraft, M. Fularz, A. Schmidt Collaborative, Context Based Activity Control Method...

Page 18: Collaborative, Context Based Activity Control Method for Camera Networks

IntroductionConcept and implementation

EvaluationConclusions and future work

Hardware platformTest setupResults

Activation levels over time (’1’ – performance, ’0’ –powersave):

M. Kraft, M. Fularz, A. Schmidt Collaborative, Context Based Activity Control Method...

Page 19: Collaborative, Context Based Activity Control Method for Camera Networks

IntroductionConcept and implementation

EvaluationConclusions and future work

Hardware platformTest setupResults

Duration of performance and powersave modes (in seconds) for thepresented test scenario

camera performance mode powersave mode % of perf. mode

PiCam01 112 249 31.02

PiCam02 140 221 38.78

PiCam03 149 212 41.27

PiCam04 77 284 21.33

PiCam05 59 302 16.34

M. Kraft, M. Fularz, A. Schmidt Collaborative, Context Based Activity Control Method...

Page 20: Collaborative, Context Based Activity Control Method for Camera Networks

IntroductionConcept and implementation

EvaluationConclusions and future work

Hardware platformTest setupResults

So, what do we get from it?

As an example, for PiCam03 (worst case) and Picam05 (bestcase) the power consumption w.r.t. the full activity mode by isreduced 12% and 17%, respectively.Please keep in mind, that the Raspberry Pi is not particularlypower efficient.Far less data is transmitted.Gives easy means of extracting and presenting the images fromthe cameras where the action takes place.

M. Kraft, M. Fularz, A. Schmidt Collaborative, Context Based Activity Control Method...

Page 21: Collaborative, Context Based Activity Control Method for Camera Networks

IntroductionConcept and implementation

EvaluationConclusions and future work

Conclusions and future workThe presented solution can successfully keep track of themovement of objects across an environment surveyed by amulti-camera system.The solution is capable of reducing the power consumption oflarge-scale, automatic surveillance systems withoutcompromising the accuracy and efficiency in terms ofmovement detection.As many advanced surveillance systems rely on backgroundsubtraction, the presented solution may be an easilyapplicable, drop-in extension of their capabilities.Great potential for future extensions – automatic gainadaptation, integration of other activity indicators...Available at https://github.com/sepherro/cam_network

M. Kraft, M. Fularz, A. Schmidt Collaborative, Context Based Activity Control Method...

Page 22: Collaborative, Context Based Activity Control Method for Camera Networks

IntroductionConcept and implementation

EvaluationConclusions and future work

Thank you for your attention(Questions? Comments?)

The project was financed by the National Science Center under the contract decision numberDEC-2011/03/N/ST6/03022,

New concept of the network of smart cameras with enhanced autonomy for automatic surveillancesystems

M. Kraft, M. Fularz, A. Schmidt Collaborative, Context Based Activity Control Method...