RAJIV GANDHI PROUDYOGIKI VISHWAVIDYALAYA,

BHOPAL New Scheme Based On AICTE

Flexible Curricula Computer Science and Engineering,

Scan for YouTube

UNIT lV

Explain Facet Model Recognition: Labeling lines, Understandin

Understandingg line drawings, Classification of shapes
by labeling of edges

Facet model recognition is a technique used to recognize and classify 2D line drawings or
sketches into different shapes based on the labeling of edges. This technique involves several
steps, including labeling lines, understanding line drawings, and classifying shapes based on the
labeling of edges.

1. Labeling lines:: In the first step of facet model recognition, lines in the drawing are labeled
based on their geometric properties, such as length
length,, orientation, and curvature. These labels
provide information about the shape and structure of the lines and are used to group similar
lines together.
2. Understanding line drawings:: In the second step of facet model recognition, the labeled lines
are analyzed
ed to understand the underlying structure of the drawing. This involves identifying the
vertices and edges of the drawing, as well as any symmetries or regularities in the shape.
3. Classification of shapes by labeling of edgesedges: In the final step of facet model
el recognition, the
labeled lines and the underlying structure of the drawing are used to classify the shape into a
predefined set of categories. This is done by matching the labeled lines and the structural
information of the drawing with a database of kn known shapes.

The labeling of lines is an important step in facet model recognition, as it provides a basis for
understanding the underlying structure of the drawing. Different labeling schemes can be used
depending on the application, such as using angles o orr curvature to label lines. Understanding
the line drawing involves identifying the vertices and edges of the drawing and extracting any
relevant features, such as symmetry or regularity. Finally, the classification of shapes based on
the labeling of edges involves matching the labeled lines and structural information of the
drawing with a database of known shapes. This can be done using various techniques, such as
machine learning algorithms or rule
rule-based systems.

In summary, facet model recognition is a ttechnique

echnique used to recognize and classify 2D line
drawings or sketches into different shapes based on the labeling of edges. This technique
involves several steps, including labeling lines, understanding line drawings, and classifying
shapes based on the labeling
eling of edges. Facet model recognition can be used in various

YouTube Channel: RGPV

PV Exam official 2023 https://www.youtube.com/@RGPVExamOfficial
 applications, such as in computer-aided design (CAD) or in the recognition of hand-drawn
sketches.

Explain Recognition of shapes, Consisting labeling problem, Back-tracking AlgorithmPerspective

Projective geometry

Shape recognition is the process of identifying the shapes or objects present in an image. The
recognition of shapes involves several steps, including the segmentation of the image, feature
extraction, and classification of the shapes. One of the main challenges in shape recognition is
the problem of shape labeling.

1. Consisting labeling problem: The shape labeling problem refers to the task of assigning labels to
the shapes in an image based on their properties. For example, a square can be labeled as a
rectangle, but not all rectangles can be labeled as squares. The shape labeling problem can be
solved using various techniques, such as graph-based approaches or statistical methods.
2. Backtracking algorithm: One technique used to solve the shape labeling problem is the
backtracking algorithm. This algorithm involves searching for the best labeling solution by
iteratively testing different combinations of labels until a valid solution is found. The
backtracking algorithm can be used for shape labeling in both 2D and 3D images.
3. Perspective projective geometry: Another important aspect of shape recognition is the use of
perspective projective geometry. This involves modeling the 3D world as a 2D projection and
using geometric transformations to map points in the image to their corresponding points in
the 3D world. Perspective projective geometry is essential for recognizing shapes in images
captured from different viewpoints.

In summary, the recognition of shapes involves several steps, including segmentation, feature
extraction, and classification. The shape labeling problem is a key challenge in shape
recognition, and techniques such as the backtracking algorithm can be used to solve this
problem. Perspective projective geometry is also important for recognizing shapes in images
captured from different viewpoints. Shape recognition has many applications, such as in
robotics, autonomous vehicles, and image analysis.

Explain Inverse perspective Projection, Photogrammetric -from 2D to 3D

Inverse perspective projection is a technique used in photogrammetry to create 3D models from

2D images. This technique involves using geometric transformations to map points in the image
to their corresponding 3D coordinates in space. The inverse perspective projection is often used
in computer vision and robotics applications to estimate the position and orientation of objects
in 3D space.

The photogrammetric process involves taking multiple images of an object from different
viewpoints and using the inverse perspective projection technique to reconstruct a 3D model.
The process begins with the calibration of the camera, which involves determining the intrinsic

YouTube Channel: RGPV Exam official 2023 https://www.youtube.com/@RGPVExamOfficial

 and extrinsic parameters of the camera. The intrinsic parameters include the focal length, sensor
size, and distortion coefficients, while the extrinsic parameters include the position and
orientation of the camera in space.

Once the camera is calibrated, the images are processed using the inverse perspective
projection technique to map 2D image points to their corresponding 3D coordinates in space.
This involves first determining the position and orientation of the camera relative to the object,
and then using the camera projection matrix to transform the 2D image points into 3D
coordinates. The resulting 3D points can then be used to reconstruct a 3D model of the object.

Inverse perspective projection is a powerful technique for creating 3D models from 2D images,
but it has some limitations. One of the main challenges is dealing with occlusions, where parts of
the object are hidden from view in some images. This can result in incomplete or inaccurate 3D
models. Additionally, the accuracy of the 3D model depends on the quality of the camera
calibration and the accuracy of the inverse perspective projection algorithm.

In summary, inverse perspective projection is a technique used in photogrammetry to create 3D

models from 2D images. This technique involves using geometric transformations to map points
in the image to their corresponding 3D coordinates in space. The resulting 3D points can then
be used to reconstruct a 3D model of the object. Inverse perspective projection has many
applications, including in robotics, computer vision, and virtual reality.

Explain Image matching: Intensity matching of ID signals, Matching of 2D image, Hierarchical image
matching
Image matching is the process of comparing two or more images to determine if they represent
the same object or scene. Image matching is used in a variety of applications, including object
recognition, image retrieval, and motion tracking.

1. Intensity matching of ID signals: One technique used in image matching is intensity matching of
ID signals. This involves comparing the intensity patterns of an object in different images and
using these patterns to identify the object. This technique is particularly useful when matching
images of objects with distinctive features, such as facial recognition.
2. Matching of 2D image: Another technique used in image matching is matching of 2D images.
This involves comparing the pixel values of two images to determine if they represent the same
scene or object. This technique is particularly useful when matching images of objects with
simple features, such as geometric shapes.
3. Hierarchical image matching: A third technique used in image matching is hierarchical image
matching. This involves breaking down the image matching problem into smaller sub-problems
and using a hierarchical approach to solve each sub-problem. This technique is particularly
useful when matching images of complex objects or scenes, such as landscapes or urban
environments.

YouTube Channel: RGPV Exam official 2023 https://www.youtube.com/@RGPVExamOfficial

 In addition to these techniques, there are many other approaches to image matching, including
feature-based matching, template matching, and machine learning-based matching. Each
approach has its own strengths and weaknesses, and the choice of technique will depend on the
specific requirements of the application.

In summary, image matching is the process of comparing two or more images to determine if
they represent the same object or scene. Techniques for image matching include intensity
matching of ID signals, matching of 2D images, and hierarchical image matching. Image
matching has many applications, including in object recognition, image retrieval, and motion
tracking.

Explain Object Models And Matching: 2D representation, Global vs. Local features
Object models and matching is an important area of computer vision that involves representing
objects in a way that allows them to be recognized and matched in images. There are many
different techniques for creating object models, but two of the most common are 2D
representation and feature-based representation.

1. 2D representation: One technique for creating object models is to use 2D representation, which
involves representing an object as a set of 2D features or landmarks. These features can be
points, lines, curves, or other shapes, and they are often selected based on their distinctive
characteristics. Once these features have been identified, they can be used to match the object
in new images by comparing their positions and shapes.
2. Global vs. Local features: Another important consideration in object modeling and matching is
the use of global versus local features. Global features are those that describe the overall shape
or appearance of an object, such as its size, orientation, or texture. Local features, on the other
hand, are those that describe specific regions or parts of an object, such as corners, edges, or
other distinctive features. Global features are often more robust to changes in viewpoint or
lighting, but they may not be as distinctive as local features. Local features, on the other hand,
are often more distinctive but may be less robust to changes in viewpoint or lighting.

Feature-based representation is another technique for creating object models. This approach
involves representing an object as a set of distinctive features or descriptors, which can be used
to match the object in new images. Feature-based representation can be more robust than 2D
representation because it is less sensitive to changes in viewpoint, lighting, or background
clutter. However, it requires more computational resources and may not be as effective for
objects with less distinctive features.

In summary, object models and matching involves representing objects in a way that allows
them to be recognized and matched in images. Two common techniques for creating object
models are 2D representation and feature-based representation. The choice of technique will
depend on the specific requirements of the application, including the complexity of the objects

YouTube Channel: RGPV Exam official 2023 https://www.youtube.com/@RGPVExamOfficial

and the robustness required for matching. The use of global versus local features is also an
important consideration in object modeling and matching, as it can affect the accuracy and
robustness of the matching process.

YouTube Channel: RGPV Exam official 2023 https://www.youtube.com/@RGPVExamOfficial