Recognition of Human Gait From Video
Marina GavrilovaComputer Science
University of Calgary
OutlineMotivationDistinguishing featuresRecognition process
Silhouette extractionHuman model initializationExtracting joint angles over image sequencesRecognition
Preliminary Results
Motivation
The goal is to detect and identify humans by the way they walk.The walking pattern (gait) is unique enough to identify a person.Such capabilities will enhance:
Human identification.Abnormal behavior detection.
Gait Cycle
Distinguishing features
Features that seem unique to each person:
Joint angle between the upper and lower legsRelationship between the knee joints and the feet over time
Elevation of knee joint over the ankle (i.e., vertical distance between knee and ankle) shows a distinctive temporal pattern
Elevation over ankle is distinctive
Transition from swingleg to stance leg is noticeably
different across different peopleover time
Gait Recognition Procedure
Image sequences
Background image
Silhouette images
Joint angles
Human model Recognition
Silhouette Extraction
Image Background
After background subtraction Final result
Human ModelHuman is modeled by five connected trapezoids. Each trapezoid (body part) is represented by
li
r1i
r2i
},,,{ 21 iiiii lrrbp θ=
iθ
Human ModelEach configuration of human body is represented by
where , and c as the center of the body.
},,,,{ 521 bpbpbpcH L=
},,,{ 21 iiiii lrrbp θ=
Human Model Initialization
2D Model-based Human tracking
Methods Cardboard person model Scaled Prismatic Model Twist and exponential maps Condensation-BasedTracking via Gibbs sampling (probabilistic)
Sample Tracking Results
Sample Tracking Results
Recognition
Collect feature vector with:Elevation over ankleJoint angles between upper and lower leg
Use left-right hidden Markov models for recognitionOne HMM per person, trained on a minimum of 4-5 full step cycles from that person
Recognition (continued)
Use algorithm similar to isolated speech recognition to identify people:
Collect a step cycle from test subjectFor each HMM in the database, compute likelihood that it matches signal of this step cycleSelect HMM with maximum likelihoodPerson corresponding to that HMM is identified subject
End of Gait Recognition
Infrared Recognition
Marina GavrilovaComputer Science
University of Calgary
#19
Primary Applications
Biometric IdentificationPasswords/PINs.Tokens (like ID cards).You can be your own password.
SurveillanceOff-the-shelf facial recognition system that identifies humans as they pass through a camera’s field of view.
#20Novel ApplicationsWearable Recognition Systems
Adapt to a specific user and be more intimately and actively involved in the user's activities. Face recognition software can help you remember the name of the person you are looking at.
Useful for Alzheimer's patients.
• Smart Systems– Key goal is to give machines perceptual abilities that allow them to
function naturally with people. – Critical for a variety of human-machine interfaces.
Why Infrared?
• Visible light has no effect on images taken in the thermal infrared spectrum.
• Even images taken in total darkness are clear in the thermal infrared.
Why Infrared? (Contd..)
Illumination InvarianceMajor problem in visible domain.
Uniqueness and RepeatabilitySense thermal patterns of blood vessels under the skin, which transport warm blood throughout the body. Remain relatively unaffected by aging. Even identical twins have different thermograms.
Immune from ForgeryDisguises can be easily detected.
Related WorkLot of research was done in the visible band but little attention was given in the infrared spectrum.Recent reduction in the cost of infrared cameras and availability of large data sets encouraged active research in infrared face recognition. Low-Level Models
Directly analyze the image pixels and impose probabilities on the features.Examples are PCA, ICA, and FDA.Not good in challenging conditions.
High-Level ModelsSynthesize images from 3D templates of known objects and imposeprobabilities on transformations.Template matching approaches.Computationally expensive.
IR Face Recognition – Training Phase
IR Face Recognition – Test Phase
Segmentation
Noise in the background may effect the performance of a face recognition system.
Remove the background.
Use thermal information on face to compute the features.
• Adaptive Fuzzy Segmentation (kakadiaris02)– Fuzzy affinity is assigned to spels w.r.t. target object spel.– Affinity is computed as weighted sum of the temperature and the
temperature gradient in the neighborhood of the target spel.– Minimal user interaction because of dynamically assigned weights.
Problem with Single Seed
Temperatures on face are different at different regions.
• If a single seed is chosen in a particular region, then the connectivity stretches only along this region and the segmentation goes wrong.
Multiple SeedsSolution to this problem is to choose multiple seeds in different regions on face and merge the resulting segmented parts .
• Choose a seed pixel on face wherever there is sharp change in gradient.
• Works well even when the subject is wearing glasses.
• Robust to variation of poses.
Choosing Multiple Seeds
Assumptions
Merge all resultant segmented regions to form final image.
ASSUMPTIONS
• The center of the image contains the pixel from facial region.• The temperatures at all pixels are mapped between 0 and 255.
– If this mapped temperature at a pixel is between 175 - 200, it is classified to be in blue region.
– If this mapped temperature at a pixel is between 200 - 225, it is classified to be in pink region.
– If this mapped temperature at a pixel is between 225 - 255, it is classified to be in cyan region.
Spectral Components
Bessel Parameters
The filtered images are modeled using Bessel parameters:
SK – Sample Kurtosis
SV – Sample Variance
• Each segmented image in training set is convolved with the filters in Gabor filter bank to obtain Gabor filtered images.
Bessel Model
Using the bessel parameters p and c, the filtered image I(j)(x,y) is modeled as:
Γ(p) is gamma function Iv(z) is modified bessel function of first kind given by:
Bessel Model
Comparing IR Images
Images modeled into Bessel parameters can be compared by:
• L2-metric between two Bessel forms f(x;p1,c1) and f(x;p2,c2) in D:
Hypothesis Pruning
Applying a high-level classifier on entire database is computationally very expensive.
Pruning of hypotheses can be achieved by using Bessel parameters (anuj01).
Helps in short listing best matches.
Bessel parameters for images in database can be computed offline which helps in saving a lot of computation time.
Hypothesis Pruning (Contd..) Shortlist the subjects of A with P1(α/I) greater than a specific threshold:
Sample Experimentswww.equinoxsensors.com
Image frame sequences were acquired at 10 frames/sec while the subject was reciting the vowels ‘a’,’e’,’i’,’o’,’u’.
End of Infrared Recognition
DNA Recognition
Marina GavrilovaComputer Science
University of Calgary
Identifiable biometric characteristics
Biological Background of DNA
Human Genome
Chromosome and DNA
DNA – The Double Helix
DNA
DNA Components
DNA Replication
Genetic Information
Characteristics of DNA
DNA in the Cell
DNA for Identification
What type of Genetic Information
Short Tandem Repeats (STR)
Short Tandem Repeats (STR)
STR Database
Comparison
Example – Two Suspects
Thank You