3.6 Signature
About1 Introduction2 Image Recognition3 TransApplet4 API5 Interface6 Input7 Image Display8 Preprocessing9 Processing10 Normalization11 Parameter Class12 Image Signatures13 Unsupervised Filters14 BioFilters15 NeuralFilters16 Dynamic Library17 NeuralNet Filter18 Parameters19 Input Options20 Database Input21 Video Input22 Live Video Input23 Counting & Tracking24 Counting 25 Batch Job26 ImageFinder for DOS27 ImageHunt 28 Support Packages
3.1 API 
3.2 Input 
3.3 Display 
3.4 Processing 
3.5 Parameters 
3.6 Signature 
3.7 Dynamic Library 
3.8 Segment Matching 
3.9 Input 
3.10 Counting 
3.11 Batch 
3.12 Customization 
[Home][3 TransApplet][3.6 Signature]

 

3.6   Signature Recognition

The Attrasoft Image Recognition Approach has two basic phases:

  •    Signature Matching

  •    Image Segment Matching

In the Signature Recognition, a unique image signature is computed for each image. Different images will have different signatures. The Signature Matching is based on these image signatures. This matching can easily reach the speed of 1,000,000 per second and is very fast.

Signature Matching is further divided into filters: Unsupervised Filter, BioFilter, and NeuralFilter. It is these filters that will perform the image matching tasks.

Chapter 12, Signature Filter, will describe how to generate image signatures.

Chapter 13, Unsupervised Filter, will describe the minimum number of steps for using the Unsupervised Filter for image recognition:

  •    Initialization

  •    Signature

  •    Unsupervised Matching

Chapter 14, BioFilter, will describe the minimum number of steps for using the BioFilter for image recognition:

  •    Initialization

  •    Signature

  •    Training

  •    Signature Matching

Initialization sets the ImageFinder parameters. Then, the image signatures are calculated and stored in a record. Unsupervised Matching can match images based on these records alone. Training teaches the BioFilter who matches with whom. After training, the BioFilter can be used for 1:1 and 1:N Matching.

BioFilter matches two whole images. BioFilter is better than Unsupervised Matching, but it requires a process called training. Training teaches the BioFilter who should match with whom. The BioFilter learns how to match the image features.

Chapter 15, NeuralFilter, introduces the Neural Filter. NeuralFilter matches two whole images, which is similar to the BioFilter. NeuralFilter is better than both the Unsupervised Filter and the BioFilter, but it requires a large amount of training data. Training data teaches the NeuralFilter who should match with whom. In comparison to early filters:

  •    The advantage of the NeuralFilter is that it is more accurate.

  •    The disadvantage of the NeuralFilter is that it requires more training data than BioFilter.

 

[Home][About][1 Introduction][2 Image Recognition][3 TransApplet][4 API][5 Interface][6 Input][7 Image Display][8 Preprocessing][9 Processing][10 Normalization][11 Parameter Class][12 Image Signatures][13 Unsupervised Filters][14 BioFilters][15 NeuralFilters][16 Dynamic Library][17 NeuralNet Filter][18 Parameters][19 Input Options][20 Database Input][21 Video Input][22 Live Video Input][23 Counting & Tracking][24 Counting ][25 Batch Job][26 ImageFinder for DOS][27 ImageHunt ][28 Support Packages]

Copyright (c) 2006 - 2007 Attrasoft. All rights reserved.

gina@attrasoft.com