Forest fires have the characteristics of suddenness, randomness, and short destruction time. Therefore, once a fire alarm occurs, it must be taken at a speed. Whether the rescue is timely or not and whether the decision is properly made depends on whether the discovery of forest fire is timely, whether the analysis is accurate and reasonable, and whether the decision-making measures are appropriate. Traditional fire alarm systems are generally based on infrared sensors and smoke sensors to detect smoke, temperature, and light parameters generated during a fire. Fire alarm signals are sent after signal processing, comparison, and judgment. The disadvantage is that smoke temperature changes caused by fire cannot be quickly collected. Information, it is difficult to meet the requirements of early detection and prediction of such fires.

       In recent years, infrared thermal imaging detection and visible light image detection have a certain degree of application in flame detection, but due to the principle of self-imaging and detection, only a single detection mode is extremely prone to false positives and false negatives, affecting user use, making this The promotion of technology has been hampered. Based on this phenomenon, the dual-spectrum detection intelligent forest fire prevention warning system uses two kinds of spectral image detection technology to maximize their respective advantages, learn from each other, effectively and accurately detect the flame, and make up for the traditional fire alarm system and single detection mode. There are deficiencies in order to achieve the effect of intelligent forest fire prevention.

       The dual-spectrum detection intelligent forest fire prevention warning system is based on the results of current forest fire monitoring technology and is based on the achievements of research and development. It is also a highly targeted system with high accuracy, high reliability and other features. Forest fire warning function.

Foreign Forest Fire Prevention Technology Development

       From the 1890s to the 1950s, temperature-sensing detectors had been dominant, and the automatic fire alarm system was in the primary development stage. In the early 1950s, the Swiss physicist Estermeyer successfully developed ionized smoke. Detectors; By the end of the 1970s, breakthroughs were made in optoelectronic components and photo-inductive smoke detectors came into being; in the early 1980s, Japan began to study experimental analog fire detectors, most notably Japanese scholars in 1991. The neural network was proposed to detect the fire source. In 1994, AlgoRex fire detection system was introduced in Switzerland. The system uses neural network and fuzzy logic to make joint decisions.

       At the end of the 1970s, some military enterprises and subordinate enterprises in China began to develop automatic fire alarm products; after entering the 1980s, in order to shorten the gap with similar foreign products and meet the needs of the domestic market, they began to import or imitate foreign products; after the 1990s, The entry of foreign companies into the Chinese market brought advanced technologies and, to a certain extent, promoted the development of the market.

       With the advancement of science and technology and the gradual upgrading of forest fire prevention information requirements, new fire detectors are also emerging; however, at present all domestic fire automatic alarm technologies are mainly based on the detection of sensors, and various existing fire alarms and Fire monitoring equipment, fire detection in most places, the use of conventional methods of fire detection, its performance will directly affect the accuracy and reliability of automatic fire alarms, such as smoke, temperature, photo detectors, they The use of flame smoke, temperature, light characteristics to detect the fire. In large-area forest applications, the above sensors are very weak due to the large spatial distance. The large space makes ordinary smoke and heat detection and fire alarm systems unable to rapidly collect information on smoke temperature changes emitted by fire. Even high-precision sensors can not work properly due to various noise interferences, resulting in false alarms or errors. Report, unable to meet the needs of timely detection of forest fires.

       At present, there are 4 billion hectares of forest in the world, of which China has 175 million hectares of forest and 1551 natural reserves. There are different levels of needs for intelligent forest fire warning systems. The best way to prevent forest fires is prevention. Countries around the world are paying more and more attention to fire warning and detection. In June 2010, the Russian fire blew up alarm bells for forest fire prevention in all countries. The lack of forest protection systems will result in forest fires or helplessness, or allow them to be left unchecked. Therefore, Russian ecology The family pointed out that the national forest protection system should be restored as soon as possible, and a practical, rapid and comprehensive early warning system is particularly important.

       The dual-spectrum detection intelligent forest fire prevention warning system adopts the dual-spectrum detection method and combines the advanced infrared thermal imager image acquisition and visible light image detection with intelligent analysis algorithms to achieve the advantages and disadvantages of each other, plus the network transmission system. The display system consists of a digital, networked intelligent forest fire prevention command system. Compared with common network video detection systems, the dual-spectrum detection forest fire prevention intelligent warning system is a higher-end, more intelligent, more accurate and reliable fire detection system that can realize unattended operation and automatically analyze and judge the video image information. , To detect the abnormal smoke and fire signs in the monitoring area in time, accurately detect the flames, and provide the fastest and best way to warn and provide fire warning information to achieve the purpose of early warning. Can effectively assist firefighters to deal with the fire crisis, and minimize the phenomenon of false positives and false negatives. At the same time, they can also view real-time images on the scene and directly command and dispatch fires according to intuitive images.

       The system can be widely used in the early warning and monitoring of forest and other outdoor fires. At the same time, combined with the forestry management expertise and experience of forestry fire prevention, a new generation of forest fire prevention intelligent analysis and monitoring system is established to solve the user's various personalized needs. Through the infrared thermal imaging and visible light dual spectrum detection, as well as the PTZ precision intelligent positioning system, a clear image of the forest area is obtained. Video analysis technology is used to determine whether a fire is generated according to the flame spectral characteristics. Once a suspected fire is found, an alarm is immediately triggered. The forest area video is transmitted back to the monitoring center. If it is confirmed that the alarm is true, the camera system locks the target, accurately determines the location of the fire point, and makes a fire extinguishing plan according to the established forest fire prevention information data source.

solution

       This system is a forest fire prevention intelligent early warning system with infrared light and visible light dual spectrum detection. The system integrates infrared thermal imager system, over temperature detection system, visible light camera, fire monitoring analyzer, PTZ precision positioning system, and video service. System, monitoring host and other components. Two video signals can be output at the same time, with infrared thermal imaging over-temperature detection and visible light fire detection capabilities, and based on the confidence factor model analysis, automatically give alarm information, effectively improve the accuracy of the alarm. The detection device can automatically scan according to any path drawn by the user in the scene, and can perform rapid pyrotechnic detection during the motion scanning process. Through the network, it transmits to the remote monitoring host information such as the alarm machine ID, the pan tilt level and pitch angle, and the coordinates of the over-temperature zone (upper left corner and lower right corner). The remote monitoring host generates alarm signals and records alarm information after analyzing and judging alarms according to the backhaul information, and provides functions such as log query and video recording.

Continuous zoom infrared thermal imager

       Continuous zoom infrared thermal imager is made of 324×256 uncooled focal plane array detector and 75-150mm continuous zoom infrared lens, which can not only search but also identify distant targets. This product overcomes the shortcomings of the current domestic or foreign fixed-focus or dual-field thermal imager. It has clear imaging, powerful functions and stable performance during zooming. With a sturdy, highly sealed housing, the interior is filled with nitrogen, protected from rain, snow and dust, and can operate in harsh environments. The 4th-generation uncooled focal plane infrared detector, the most advanced electronic and optical system in a thermal imager, can penetrate dust, smoke, rain and snow and dark, the minimum temperature resolution of 50mK, increase the image detail enhancement Function, output hot white/hot black/pseudo-color image.

Thermal imaging over temperature detection system

       In large areas of forests, fires are often triggered by hidden fires. This is the root cause of devastating fires. With the current common detection methods, it is difficult to find such hidden fire signs. The use of infrared thermal imager can quickly and effectively find these hidden fires, and can accurately determine the location and scope of the fire.

       Any object with a temperature higher than absolute zero in the natural world will continuously radiate electromagnetic waves including the infrared band to the surrounding space. The higher the temperature of the surface of the object, the more infrared radiation energy, so infrared radiation can be used to measure the surface of the object. Hot state.

    The thermal imager works at 8-14μm and belongs to the far-infrared region. The normal forest radiation wavelength range is 8.5-12.2μm, which is within the detection range of the thermal imager 8-14μm. The higher the target temperature, the greater the digital signal output from the thermal imaging detector assembly, ie the greater the gray value of the digital image. According to this characteristic, the over-temperature detection working process is as follows: the image acquisition module will output the detector with high precision The image data is written to the memory, and the image processing module runs an over-temperature detection algorithm. First, the original threshold is calculated based on the contrast between the target and the background. Combined with the target temperature level set by the user, the binarization threshold is calculated and the image is binarized. Connected area detection is performed, the area and coordinates of the target area are calculated, an over-temperature area is identified on the screen, and an alarm message is sent through the serial port.

       Because the infrared radiation energy received by the detector is affected by the monitoring distance and the working environment, the temperature range of the detected target is also different, so in order to achieve the desired alarm effect, the monitored target can be set according to the user's specific use environment. Temperature level, which is the temperature difference between the target and the background.

Visible light image detection system

       Due to the poor imaging resolution of infrared thermal imager, there may be a certain degree of false positives. Therefore, the system introduces the detection of visible light images and detects the static features (color) and morphological features (sparkle) of the flame. Firstly, the static features are used to extract the regions with similar flame colors from the video image, and then the morphological features are used to detect the above extracted regions. Through the video image analysis algorithm, it is detected that the flame generates secondary alarm signals. The visible light camera analog video signal is connected to the image detection module and converted into a digital signal by the video decoding circuit of the image acquisition unit and processed by the image processing unit based on the DSP. According to the image characteristics of the fire flame, the flame appearing on the screen is detected. After the flame identification mark is added, it is converted to an analog video signal output through a video encoding circuit. Due to the different monitoring scenes, the color and state of the fire will be different. Therefore, during monitoring, the working status of the detection module can be adjusted according to environmental requirements, and by setting corresponding parameter thresholds, such as color sensitivity, dynamic sensitivity, etc., the detection module can identify the flame more accurately and timely.