Specifications Which Allow Deployment Under A Wide-Range Of Conditions, Budgets And Mission


Built To Scale To Almost Any Wild Fire Detection Challenge:

The WDS system is enhanced with various user selected options to improve the precision of the decision-making including:

• In the event of a system application that contains camera movement, the user can select an option that has the system relearn the background image at each parked position on each revolution.
• In order to insure detection of 100% wildfire initiations while minimizing false positives or alarms the user can:
-Adjust the detection sensitivity of the system
-Extend the motion evaluation time by selecting an option that will hold the image stream evaluation in memory while the system moves to the next parked position, evaluates that parked position, then returns to again analyze the dynamics of the held video evaluation. This allows for greater expansion of the potential smoke moving region.
-Select the addition of an additional support vector machine decision making algorithm set that has been trained with a massive data set beyond basic smoke characteristics.

• To improve detection at long ranges where the moving region dynamics are minimal, the user can select an option that holds an image stream evaluation in memory for a complete rotational cycle, comparing the image streams from both cycles, reduces the threshold limit for motion detection and utilizes the additional support vector machine process from above. Specifications to accomplish the system's critical mission include:

General Specifications
Range: Detection range is dependent on camera Resolution, Field of View, and Zoom position. The system detects smoke plumes within a 10 x 10 pixel block. (for a 640 x 480 resolution camera this equals 0.03% of the image)
Area: 300 sq Km (4800 sq km with 16 Cameras)
Rotation: 360 ° with 125 set points
Sensors Types: Analog or IP CCTV cameras: Requirement includes a ¼" or larger ccd imager, glass lens and minimum 10X zoom
Sensor Capacity: 16 for WFDS-FPM and WFDS-STD and 25 for WFDS-Premium
OS Requirements: Windows : Windows XP(SP3),Windows 7, Windows Server R2008
Hardware Requirements for 16 Cameras
CPU: Quad Core, 8 thread, Hyperthreading and 8 megabytes of cache memory
Memory: 4 Gigabyte and above, recommend DDR2 or DDR3
Permanent Storage: IDE/SATA/SCSI, recommend RAID5/6, 1TB or greater
Hardware Requirements for 25 Cameras
CPU: 2 socket server populated with Intel E-5 2643 or better
Memory: 16 Gigabyte and above, recommend DDR3
Permanent Storage: IDE/SATA/SCSI, recommend RAID5/6, 1TB or greater
Supported IP Cameras
• Axis 20X, 21X, 23X, and P55 Series
• Axis Video Encoders Q24 Series and M70 Series
• JVC, Samsung and LG analog cameras have been deployed
• Customer installed camera with WDS approval
Core
• Supports both digital and analog cameras
• Supports up to 16 cameras per work station
• Adjustable frame rate provides options for smaller bandwidth capability
• Variable dwell time after movement (plume) detected

Analyzer

The analyzer is the brains of the WFDS System. It provides three key modules and multiple configurations for optimal set up in each application. There are three core detectors. The two daytime detectors detect the rising smoke and the night detector detects the growing size of the fire image:

• Color Threshold: Is set to 128-255 by default which corresponds to gray-to-white color range. Forest fires are almost always in this range during the initial stage. The settings can be modified to detect darker smoke, for example settings to to 0-255 would detect all colors, and 0-128 would detect black-to-gray colored smoke. (ie dark factory smoke)
• Sensitivity - The default level for sensitivity is 80%. Increasing the percentage will make the detector more sensitive and may increase the false positives. A decrease will make it less sensitive and also decrease false positives. An example is that a higher percentage should be used in wilderness areas and a lower percentage in areas of heavy traffic and buildings.

Analyzer Nuisance Alarm Reduction

• "Find Smoke Between Presets" - The camera saves the current image of the preset to the memory and when it returns to the preset for the second time, the saved image and current preset image are compared for the existence of smoke.
• "Return to alarm preset" - When alarm is issued for a preset, possible smoke regions are marked but the sound is not enabled. After that camera moves to the next preset, when the detection in this preset is completed, camera moves back to the previous preset to check for smoke again. If still a alarm is issued, the sounds are enabled.
• "Reset when camera shakes" - This features causes the system to reset the background image in memory due to camera vibration
• "Smoke should be brighter than background" - This is also similar to the color condition, for detecting very dark smoke, you can disable this to get better results. For wildfire detection this should be left selected.
• "Use SVM classification" - This selects and additional decision making alogrithm set that has been trained with a massive data set beyond basic smoke characteristics. It will then compare elements of the live image with those of the data set to assist in detection decision to reduce nuisance alarms

False Alarm Regions

False Alarm Region (FAR) provides the ability to have the analyzer function limited to a portion of the video stream from the camera. This allows the user to eliminate nuisance alarms that are created by vibrating structures, factory or other permanent smoke and moving clouds. "Select Excluded Regions" An option exists to automatically select potential regions to be excluded from the analysis. "Exclude Mountain Edges" When enabled this finds the edges of the mountains and adds these regions to the false alarm regions.. This settings is added to automatically find the mountain edges that cause false alarms when the camera shakes. "Exclude White Regions" : similar to "Exclude Mountain Edges", finds all the smoke colored regions and adds them to the false alarm regions. This is used to decrease false alarms issued to smoke colored regions (clouds).

Alarm

In the Alarm Settings Section, user can set an email address to send an email when alarm (Intelligent Video Alarm type must set before the Alarm Settings) occurs, set the system to stop when an alarm occurs and turn the audible alarm on and off.

Recording Settings

The WFDS system provides the capability for recording of each cameras monitoring. The recordings can be replayed by the WFDS system or any video player that supports a .avi format. The record length for each file, allocated hard drive space, alarm only recording, and web image transfer are all user selectable. A text file (.txt) is also created for each alarm or warning generated and stored in the folder with the video clips. The default location is the primary WDS program file

Manual Control

When an alarm occurs, the operator can take control of the camera, examine the event further and determine the coordinates of the event to dispatch suppression crews.

Remote Monitoring

The WFDS system is equipped to be monitored and controlled from remote sites. In fact, multiple sites can monitor the system. Any of these sites can also control the system.