2. Object recognition in computer vision is the task of finding a given object in an
image or video sequence. Humans recognize a multitude of objects in images
with little effort, despite the fact that the image of the objects may vary
somewhat in different view points, in many different sizes / scale or even when
they are translated or rotated. Objects can even be recognized when they are
partially obstructed from view. This task is still a challenge for computer vision
systems in general.
Object recognition concerned with determining the identity of an object being
observed in the image from a set of known labels. Oftentimes, it is assumed that
the object being observed has been detected or there is a single object in the
image.
Object recognition system finds objects in the real world from an image of the
world from an image of the world, using object models which are known a priori.
Humans perform object recognition effortlessly and instantaneously
3.
4. An object recognition system must have the
following components to perform the task:
Model Data Base
Feature Detector
Hypothesizer
Hypothesis verifier
5. Model Data Base - contains all the models known to the system. The
information in the model database on the approach used for recognition. The
models of objects are abstract feature vectors, as discussed later in this section.
A feature is some attribute of the object . Size, color, and shape are the
commonly used features.
Feature Detector – applies operators to images and identifies locations of
features that help in forming object hypothesis. The features used by a system
depend on the types of objects to be recognized.
Model Data Base – Using the detected features in the image, it assigns
likelihoods to objects present in the scene. Used to reduce the search space for
the recognizer using certain features.
Verifier– uses object models to verify the hypotheses and refines the
likelihood of objects. The system then selects the object with the highest
likelihood, based on all the evidence, as the correct object.
6. Object or model representation: How shpuld objects be represented in the
model database? – For some objects, geometric descriptions may be available and
may also be efficient, while for another class one may have to rely on generic or
funtional features.
Feature Extraction: Which features should be detected, and how can they be
detected reliably? – Most features can be computed in two dimensional images but
they are related to three-dimensional characteristics of objects.
Feature-model matching: How can a set of likely objects based on the feature
matching be selected? – this step uses knowledge of the application domain to
assign some king of probability or confidence measure to different objects in the
domain.
Object Verification - How can object models be used to select the most likely
object from the set of probable objects in a given image? – The presence of each
likely object can be verified by using their models.
7. Scene Constancy: the scene complexity will depend on
whether the mages are acquired in similar conditions
(illumination, background, camera parameters, and
viewpoint) as the models.
image-models spaces: Images may be obtained such that
three-dimensional objects can be considered two-
dimensional.
Number of Objects in the model database: If the number
of objects is very small, one may not need the hypothesis
formation stage.
Number of objects in an image and possibility of
occlusion: If there is only one object in an image, it may be
completely visible.
8. Two-Dimensional
In many applications, images are acquired from a distance sufficient to
consider the projection to be orthographic. If the objects are always in one
stable position in the scene, then they can be considered two-dimensional.
In these applications, one can use a two-dimensional model base. There are
two possible cases:
Objects will not be ocluded, as in remote sensing and many industrial
applications.
Objects may be occluded by other objects of interest or be partially
visible, as in the bin of parts problem
9. Three-Dimensional
If the images of objects can be obtained from arbitrary viewpoints, then an
object may appear very different in its two views. For object recognition using
three-dimensional models, the perspective effect and viewpoint of the image
have to be considered. The fact that the models are three-dimensional and the
images contain only two-dimensional information affects object recognition
approaches. Again, the two factors to be considered are whether objects are
separated from other objects or not.
For tree-dimensional cases, one should consider the information used in the object
information used in object recognition task. Two different cases are:
Intensity: There is no surface information available explicitly in intensity
images. Using intensity values, features corresponding to the three-dimensional
structure of objects should be recognized.
2.5-dimensional images: In many applications, surface representations with
viewer-centered coordinates are available, or can be computed, from images.
This information can be used in object recognition.
10. 3D object recognition based on the use colored stripes—so called structured light—is
useful in applications ranging from 3D face recognition to measuring suspension
systems and ensuring a perfect fit for hearing aids
11.
12. Uses description of objects in a coordinate system attached to objects. This description
is usually based on three dimensional features or description of objects. These are
independent of the camera parameters and location. Thus, to make them useful for
object recognition, the representation should have enough information to produce
object images or object features in images for a known camera and viewpoint.
a.) an object is shown with its prominent
local features highlighted.
b.) graph representation is used for object
recognition using a graph matching
approach.
13. Many types of features are used for object recognition. Most features are based on
either regions or boundaries in an image. It is assumed that a region or a closed
boundary corresponds to an entity that is either an object or a part of an object.
An object and its partial
representation using multiple
local and global features
14. Depending on the complexity of the problem, a recognition strategy
may need to use either or both the hypothesis formation and
verification steps
15.
16. Face recognition is a rapidly growing field today for is many uses in the fields
of biometric authentication, security, and many other areas. There are many
problems that exist due to the many factors that can affect the photos. When
processing images one must take into account the variations in light, image quality,
the persons pose and facial expressions along with others. In order to successfully be
able to identify individuals correctly there must be some way to account for all
these variations and be able to come up with a valid answer.
Figure
Differences in Lighting and Facial Expression
17. Face recognition is an image processing application for automatically
identifying or verifying a person from a digital image or video frame from a video
source. One of the ways to do this is by comparing selected facial features from the
image and a facial database.
Some facial recognition algorithms identify faces by extracting landmarks, or
features, from an image of the subject's face. For example, an algorithm may
analyze the relative position, size, and/or shape of the eyes, nose, cheekbones,
and jaw. These features are then used to search for other images with
matching features.
Other algorithms normalize a gallery of face images and then compress the
face data, only saving the data in the image that is useful for face detection.
18. Face recognition used in:
- Human and computer Interface
- Biometric identification
Objective of Face recognition :
-to determine the identity of a person from a given image.
Complications occur due to variations in:
- Illumination
- Pose
-Facial expression
-Aging
-occlusions such as spectacles, hair, etc.
Weaknesses:
-Face recognition is not perfect and struggles to perform under certain
conditions.
-Other conditions where face recognition does not work well include poor
lighting, sunglasses, long hair, or other objects partially covering the subject’s
face, and low resolution images.
-less effective if facial expressions vary
19. Facial Recognition uses mainly the following techniques:
•Facial geometry: uses geometrical characteristics of the face. May use several
cameras to get better accuracy (2D, 3D...)
•Skin pattern recognition (Visual Skin Print)
•Facial thermogram: uses an infrared camera to map the face temperatures
•Smile: recognition of the wrinkle changes when smiling
20. The uniqueness of Skin Texture
offers an opportunity to identify
differences between identical
twins.
The Surface Texture Analysis
algorithm operates on the top
percentage of results as
determined by the Local feature
analysis.
21.
22. Finger Printing is one of the most well-known and publicized biometrics. Because of
their uniqueness and consistency over time, fingerprints have been used for
identification for over a century, more recently becoming automated due to
advancements in computing capabilities. Fingerprint identification is important
because of the inherent ease in acquisition, the numerous sources available for
collection, and their established use and collections by law and immigration.
A Fingerprint usually appears as a series of dark lines that represent the high,
peaking portion of friction ridge skin, while the valleys between these ridges appears
as white space and are low, shallow portion of the friction ridge skin.
Finger Identification is based primarily on the minutiae, or the location and
direction of the ridge endings and bifurcations (splits) along a ridge path
23. Hardware
A variety of sensor types – optical, capacitive, ultrasound, and thermal – are
used for collecting the digital image of a fingerprint surface.
•Optical sensors take an image of the fingerprint, and are the most common
sensor today
•Capacitive sensor determines each pixel value based on the capacitance
measured, made possible because an area of air has significantly less
capacitance than an area of finger.
Software
The two main categories of fingerprint matching techniques are minutiae-based
matching and pattern matching.
• Pattern Matching simply compares two images to see how similar they are.
Usually used in fingerprint systems to detect duplicates.
• Minutiae-based matching relies on the minutiae points described above,
specifically the location and direction of each point.
24. Geometry-based approaches- early attempts on object recognition
were focused on using geometric models of objects to account for their
appearance variation due to viewpoint and illumination change.
Appearance-based algorithms- advanced feature descriptors and
pattern recognition algorithms are developed. Computes eigenvectors from a set
of vectors where each one represents one face image
Feature-based algorithms- lies in finding interest points, often occured
at intensity discontinuity, that are invariant to change due to scale, illumination
and affine transformation.