CCTV and imaging technology

Training and Test Datasets both contain 20-24 hours of footage for each Scenario. The Training Dataset splits the footage into smaller sequences, normally 3 - 6 minutes. The Test Dataset has longer duration sequences, normally 30 - 60 minutes. VBDS should be trained on the Training Dataset to optimise the VBDS performance. VBDS can then be tested on the Test Dataset. Performance results from the Test Dataset can then be submitted to HOSDB.

No. VBDS developers are not required to obtain both datasets, Training and Test Datasets can be ordered individually. For a system to be considered for evaluation, testing on the relevant Test Dataset must have been conducted by the manufacturer.

The image sequences (or "clips" as they are known) are captured in MJPEG format CIF-4 resolution of 576*704 (4:3 aspect ratio) with 25 interlaced frames per second and 8 bit colour quantisation. Individual frames are compressed to 90% of their original size. As this is MJPEG, there is no interframe dependency.

Submit your F1 Result from the Test Dataset of the relevant scenario along with the Evaluation Application Form, which can be found on this site.

VBDS act as a trigger for recording of suspicious events, where all the recordings obtained are to be analysed at a later time.

VBDS provide real-time detection of suspicious events which must be dealt with by a human controller.

Check the relevant box on the Dataset Application Form and our Distributor will contact you to arrange other methods of payment.

Guidance is available in the i-LIDS User Guide, which can be found on this site.

Yes. Email i-LIDS enquiries with your name, address and company details and we will send you a demonstration DVD.