|
Since the introduction of video surveillance as a security tool there has existed a twofold problem. The first is inattentional or change blindness. That is observers do not reliably detect security threats, when watching video. As a result the conclusion is drawn, that nothing has occurred when, in fact, something has happend. A Harvard experiment has shown that, while watching video, 50% of observers counting the passes made between two basketball teams will not notice a gorilla walk into the middle of the viewing area, beat its chest, and walk out. Similarly, in a military study, experiments demonstrated that after 12 minutes of continuous viewing of 2 or more sequencing monitors, an operator will miss up to 45% of all scene activity. After approximately 22 minutes, an operator will miss up to 95% of scene activity.
The second issue is visual acuity. Shown in figure 1 is a typical human eye scan path. Fixations are shown by circles and scans by red dashed line. The number of scans and fixations, and thus the time to detection, is a function of the number of confusing objects (potential points of interest ‘fixations’). When the eye has locked onto the target as shown in figure 2, the entire image in the peripheral vision is severely degraded. It has been shown that the probability of detection is a direct function of resolution. As a result if the operator is studying a point, such as the climber in the photo, he is less likely to detect another intruder in the image.
This problem of peripheral vision is further aggravated in that the observers are often required to monitor more than one screen to detect security threats (see Figure 3).
No matter how highly trained or how dedicated a human observer, it is impossible to provide full attention to more than one or two details at a time; and even then, only for a few minutes at a time.
The true measure of a video surveillance system effectiveness is the level of responsiveness to security breaches. If this response can be enhanced, then there is a significant increase in the value of the system. This suggest the requirement for a simple, real time intelligent video, system to stimulate (rouse) the operator and identify the search point to minimize the detection time. The Real Time Change Detector (RTCD) (patent pending) provides just such an enhancement. The fractional cost of the RTCD will provide multiples of system performance.
The RTCD concept is simple and straight forward, utilizing a powerful computer. The human mind processing the images from the eye is one of the most powerful computers known. It can interpolate data based on history. It can reject false alarms based on similar patterns. It can identify targets based on patterns and changes in patterns. It can interpolate missing pieces of data. The problem is that the mind gets bored when there are none of the above functions to perform.
The Real Time Change Detector (RTCD) rouses the bored mind of the operator by visually identifying that a change has taken place. It then utilizes this robust computing capability of the observer to identify the specific location on the correct display, so that he can rapidly focus his concentration (foveal vision) on the potential intruder.
Continued on next page
|
