Scribd
Upload
7 views13 pages

Identity Declaration Pattern Recognition

Uploaded by

Ratheesh P M
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
7 views13 pages

Identity Declaration Pattern Recognition

Uploaded by

Ratheesh P M
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 13

Identity Declaration & Pattern Recognition

Nithin K Mohan |2nd Semester | 2021-2023 | M. Tech Instrumentation Technology | DOI | CUSAT |
Identity Declaration & Pattern Recognition

Nithin K Mohan |2nd Semester | 2021-2023 | M. Tech Instrumentation Technology | DOI | CUSAT |
Identity Declaration & Pattern Recognition

Identity declaration involves transforming observed attributes of an entity


such as size, shape, and spectral characteristics into a label that describes
or names the identity of the entity.

Techniques for identity declaration are drawn from the literature of pattern
recognition

Nithin K Mohan |2nd Semester | 2021-2023 | M. Tech Instrumentation Technology | DOI | CUSAT |
Identity Declaration & Pattern Recognition

Nithin K Mohan |2nd Semester | 2021-2023 | M. Tech Instrumentation Technology | DOI | CUSAT |
Identity Declaration & Pattern Recognition

The role of identity fusion within level 1 processing

Nithin K Mohan |2nd Semester | 2021-2023 | M. Tech Instrumentation Technology | DOI | CUSAT |
Identity Declaration & Pattern Recognition
Features are a representation of the data that may be useful in the identification
process.

Concept of single sensor identity declaration .

Nithin K Mohan |2nd Semester | 2021-2023 | M. Tech Instrumentation Technology | DOI | CUSAT |
Identity Declaration & Pattern Recognition

The identity declaration process uses pattern recognition techniques such as

• Templating
• Cluster analysis
• Adaptive neural networks
• Knowledge-based techniques to recognize an entity's identity.

Nithin K Mohan |2nd Semester | 2021-2023 | M. Tech Instrumentation Technology | DOI | CUSAT |
Identity Declaration & Pattern Recognition

Each of these techniques involves a two-phase operation.

In the first phase (the training phase), data from known entities are used to establish
the boundaries that partition the feature space into regions, each region establishing
identity of an object or entity (i.e., the parametric boundaries for templates or the
cluster boundaries for cluster analysis).

In the second phase of operation, having established the relationship between


feature vectors and identity, the pattern recognition techniques are used in a
recognition or classification mode. In such a mode, individual feature vectors are
classified by establishing the location of the observed vector with respect to the
decision boundaries in feature
space.

Nithin K Mohan |2nd Semester | 2021-2023 | M. Tech Instrumentation Technology | DOI | CUSAT |
Identity Declaration & Pattern Recognition

Nithin K Mohan |2nd Semester | 2021-2023 | M. Tech Instrumentation Technology | DOI | CUSAT |
Identity Declaration & Pattern Recognition

Decision-level approach
Each sensor performs independent processing to produce an estimate, and
estimates are subsequently combined via a fusion process. Techniques for fusing
declarations of identity include voting methods , Bayesian inference, Dempster-
Shaffer’s method, generalized evidence processing theory , and ad hoc method

Nithin K Mohan |2nd Semester | 2021-2023 | M. Tech Instrumentation Technology | DOI | CUSAT |
Identity Declaration & Pattern Recognition

Feature-level fusion approach

In this approach, each sensor observes an object, and a feature extraction is


performed to yield a feature vector from each sensor. These feature vectors are
then fused and an identity declaration is made based on the joint feature vector

Nithin K Mohan |2nd Semester | 2021-2023 | M. Tech Instrumentation Technology | DOI | CUSAT |
Identity Declaration & Pattern Recognition
Data Level fusion approach

Data from commensurate sensors are fused directly, with subsequent feature extraction and identity
declaration from the fused data. In order to perform such data-level fusion the sensors must either
be identical (e.g., several IR sensors), or commensurate (e.g., an IR sensor an visual imagery
sensor). Association is performed on the raw data to ensure that data being fused relates to the
same object or entity. Having fused the sensor data, the identification process proceeds identically to
the process for a single sensor

Nithin K Mohan |2nd Semester | 2021-2023 | M. Tech Instrumentation Technology | DOI | CUSAT |
Thank you

Nithin K Mohan |2nd Semester | 2021-2023 | M. Tech Instrumentation Technology | DOI | CUSAT |

You might also like