SIVLI - FAQ.

Frequently Asked Questions.

1. What does SIVLI mean?
Intelligent Software for Video Location of Images.

2. How does SIVLI work?
SIVLI is a patenting-in-process computer system that is the result of a combination of state-of-the-art techniques that are part of the following areas:

• Computational Intelligence.
• Computer Vision.
• Signal Processing.
• Computational Geometry.

Given input image and video sequence, SIVLI performs an analysis over the frames composing the sequence and identifies those with a high similarity degree with the input image.

The frames with the highest similarity degree are extracted and condensed in a set which is returned as output.

The notion of Video Location takes place by annexing to each selected frame, in the output set, its corresponding position in the video sequence and its Similarity Value.

The Similarity Value in each frame in the output set shared by SIVLI is a number taking values between 0 and 1. If the Similarity Value between a frame and the input image is:

• 0, then frame and input image must be characterized as identical.
• 1, then frame and input image must be characterized as completely different.

If the Similarity Value between a frame and the input image is enough close to 0 then we have frames with common elements but also with some differences.

SIVLI only returns as output those frames with the best Similarity Values.

From the User’s point of view, SIVLI only receives as input an image and the video sequence over which the search is going to be performed.

In consequence, SIVLI returns as output the corresponding set of frames with the best Similarity Values.

3. What are the advantages shared by SIVLI to the User?

• Wide range of image formats:
• Image formats accepted by SIVLI are those popularly supported and generated by digital video cameras, cell phones and tablets (JPG, GIF, BMP, PNG, etc.).
• Wide range of video formats:
• Video formats accepted by SIVLI include, among others, those popularly supported and generated by digital video cameras (AVI, MPEG-1, MPEG-2, MPEG-4, WMV, QuickTime, etc.), cell phones and tablets (MP4, 3GPP, 3GPP-2, etc.), and also those related to streaming technologies (FLV, NSV, etc.).
• Independence over image and video sequence resolutions:
• In principle, SIVLI is capable of processing images and video sequences of any resolution. However, a minimal acceptable resolution that allows useful analysis results is the corresponding to Analogical Television: 320 x 240 pixels.
• Independence over video sequence contents:
• SIVLI rises as a versatile tool which performs analysis and recognition specific and adequate to each video sequence and image shared as input.
• This implies there is no problem for processing video sequences whose content corresponds to surveillance (with static or moving cameras), motion pictures, television programs, streaming, etc.

4. How much time is required for development?
At the moment, SIVLI is in a beta tester version. It is in development an alternative version whose main goal is the sharing of results in a more efficient way.

5. Who should be contacted for more information?
Ricardo Pérez-Aguila, PhD.
ricardo.perez.aguila@gmail.com
http://ricardo.perez.aguila.googlepages.com

Ricardo Ruiz Rodríguez, MSc.
ricardo.ruizrodriguez@gmail.com
https://sites.google.com/site/ricardoruizrodriguez/