Wildland Systems’ Parent Company Receives Critical Video Smoke/Vapor Detection Patent

Delacom Detection Systems, LLC, (DDS) parent of automated wildfire monitoring/detection supplier Wildland Systems, has been granted a US patent on a new form of video detection technology. This new approach is being used in all of the company’s technologies to provide low-cost, high reliability automated detection of smoke, vapors and even unwanted intruders indoors or out in the field.

Reduces the Number of Needed Fire Sensors

The technology employs a method called wavelet transformation to detect differences between what is considered a “normal” background image and an image that is “abnormal” such as smoke, vapor leaks or even the outline of a vehicle or person that may appear as part of the sampled image. By being able to use either visible or invisible light (such as infrared) captured by video cameras, “aberrations” such as leaks or smoke can be detected more accurately with a single point of collection which lowers the cost over employing separate sensors. It also allows a tremendous increase in range over these discrete sensing devices (such as portable flame ionization detectors in the case of vapor detection) as a single video camera may actually be able to monitor up to 300 square kilometers. This is tremendously useful in situations like monitoring pipelines, tank farms, and of course open landscapes. It can replace the complex networks required to support large numbers of discrete sensors.

The invention makes use of the properties of wavelet transformation which can be finitely controlled to reduce the signal to noise levels of the image, making the elements which need to be detected stand out more clearly for later analysis. While wavelet transformation is not a new technology per se (and has been widely employed in applications such as video image compression/decompression to increase signal transmission efficiency), the innovation granted to DDS deals specifically with applying a series of analytical techniques to the wavelet-generated signal which DDS categorizes as “adaptive” analysis versus current “non-adaptive” interpretation used by other approaches.

Volatie Organic Compound

Type

Number
of
Frames
with
Visible
Plume

Number of
Detection-
Ready
Frames

Number of 
False
Positives

 

 

Adaptive Thresholds

Non-Adaptive

Adaptive Thresholds

Non-Adaptive

Gasoline

1241

1120

1088

0

0

Diesel

443

405

265

0

38

Propane

310

288

120

0

14

Results for the older style of non-adaptive or fixed versus DDS new adaptive threshold detection methods for different volatile organic compounds (VOC) types are shown above. The wavelet-based detection method which was just granted a full US patent, shows a higher detection performance/ accuracy even for semitransparent VOC plumes over older interpretation approaches. This study cited in the patent application notes virtually no false alarms are issued for regular moving regions such as people, cars, etc. when adaptive thresholds are used. In addition, the computing required for signal processing is low which enables real time detection on standard personal computing networks. The processing time per frame is less than 15 milliseconds for 320 by 240 pixel frames.

Develops New Model for Detecting Smoke and Vapor

These techniques include a method of “subtracting” the background of an image from the monitored scene and analyzing  patterns associated with vapors or heat disturbances.  Another approach (based on Markov’s principles), examines the intensity of the various light detected elements making up the scene. In effect, this approach uses light detection as a means for detecting the energy levels output by varying substances that are being observed which the wavelet transformation represents in different intensities.  Part of the innovation is also in the weighting of the value of each of these elements in the model it creates to determine vapor/smoke presence via the patterns being recognized.

The approach was invented by a team of scientists under the direction of Professor A. Enis Cetin, Chief Technology Officer (CTO), Wildland Detection Systems. While the basic technology has been employed in WDS and DDS products since it was applied for in 2008, the granting of a full patent by the United States Patent Office provides full legitimacy of this ground-breaking approach, and underscores the improvement in leak/smoke/pattern detection that it brings to the world of detection/monitoring of wildlands, pipelines, tank farms, cargo ships and containers and more.

Dennis Akers, CEO & President of DDS recently observed that “many industries need the assurance of being able to detect and prevent the calamities that can occur when smoke, vapor or other precursors are not detected in time to stop fire and explosions. This technology can now do this more cheaply and more dependably than other methods.”  Akers notes that the proof of the value of this technology is that it has been operating world-wide with over 150 installations on three continents, collecting over 300,000 continuous hours of monitoring and has yet to fail at detecting a fire while being operational. He invites all who are interested to review this important technology and contact his company to discuss your specific situations and needs.