CN103136538A - Multidata fusion flame recognition device and method - Google Patents

Abstract

The invention provides a multidata fusion flame recognition device and method. After the device and device are applied to the internet of things, the problem that node electric quantity consumption is too fast is well solved. A data fusion and pattern recognition algorithm is added, fire prediction success rate is improved, and on the contrary, energy consumption of a node is reduced. The device is effectively combined with a network in a zigbee type in a wireless sensor network (WSN), so long-distance wireless transmission of trace amount of valid data is achieved successfully. Especially after image recognition technology is applied to the device, requirements of staff for forest fire prevention are reduced, automatic collection of valid data in a wireless network is achieved, surplus data are eliminated, problems inherent in hardware in the WSN are well made up, and the device can effectively collect and feed back information in the internet of things.

Description

A multi-data fusion flame recognition device and method

technical field

The invention belongs to the technical fields of Internet of Things, forest fire prevention, etc., and relates to a multi-data fusion flame recognition device and method.

Background technique

The Internet of Things acquires data in real time and collects environmental information, which has been applied in many fields. With the support of information technology, the Internet of Things is triggering a new round of lifestyle changes. With the development of the Internet of Things, a series of scientific research problems have also emerged one after another. The author of this paper found such problems in the process of applying the Internet of Things technology to forest fire prevention. For example, due to the large power consumption of WSN and the short network life, the availability of nodes is relatively poor. Due to the single detection capability of the nodes in the WSN of the Internet of Things, data confusion is caused.

Contents of the invention

The purpose of the present invention is to overcome the above-mentioned technical defects and provide a multi-data fusion flame recognition device and method.

Its technical solution is:

A multi-data fusion flame recognition device, including a sensing module, a communication module, an image acquisition module, and an arduino single-chip microcomputer, wherein the single-chip microcomputer includes a processing module, a data acquisition module, and a data fusion module:

Sensing module: collect basic data through temperature and humidity sensors, gas sensors, fire detection sensors and image sensors for data fusion;

Processing module: complete the data fusion calculation and image recognition algorithm through the Arduino single-chip computer, so as to judge the valid data and then pass it to the communication module. At the same time, the single-chip computer also has the function of power management to provide power strategy for the whole device;

Communication module: Xbee protocol stack implemented by hardware;

Data acquisition module: it is the driver program of the sensor, which drives the sensor to collect data, and automatically eliminates redundant, invalid, and mixed sensor data through the data gate value, so as to achieve the effect of power consumption;

Image acquisition module: used to acquire images, and then perform pattern recognition, which integrates the video driver and pattern recognition algorithms on Arduino;

Data Fusion Module: Combine all kinds of information into key information and report possible fires.

The temperature and humidity sensor is a DHT11 digital temperature and humidity sensor with calibrated digital signal output.

The gas sensor is based on the MQ3 gas sensor of the gas sensor, which is used to detect alcohol and ethanol gas in the air, and is used in combination with the Arduino special sensor expansion board to detect the concentration of smoke in the air.

The fire source detection sensor can detect light sources with wavelengths in the range of 760 nm to 1100 nm, convert them into digital signals and transmit them to the processing module.

The image sensor CMUcam4 is a programmable embedded image sensor, which uses Parallax P8X32A chip and Omini Vision9665CMOS camera module to capture photos of flames in the forest.

A multi-data fusion flame identification method, a multi-data fusion flame identification method, first collects data through a camera, then uses a single-chip microcomputer to delete valid data, deletes image information through a gate value, and finally synthesizes valid graphic data, To identify the flame.

Compared with prior art, the beneficial effect of the present invention is:

After the device is applied to the Internet of Things for forest fire prevention, it solves the problem of high power consumption of nodes. The data collected by the Internet of Things can more accurately and directly reflect the changes of various environmental factors at the monitoring site, and has better timeliness, which can well make up for the shortcomings of traditional fire monitoring and early warning technologies. The project uses data fusion technology to solve the current forest fire monitoring and early warning through the tracking of sensor sensing information in the fire monitoring process and the information fusion decision-making of multiple information sources. What is even more rare is that the algorithm and software used by the author make up for the inherent problems of the hardware in WSN, and invent an effective device for collecting and feeding back information in the Internet of Things.

Fig. 1 is the hardware connection structural diagram of the flame identification device of the present invention that adorns multi-data fusion;

Fig. 2 is the software structural diagram of the flame identification device of multi-data fusion;

Figure 3 Process diagram of multiple data fusion.

Detailed ways

The device of the present invention will be further described in detail below in conjunction with the drawings and specific embodiments.

The invention constructs multiple data fusion models, arranges the obtained data in time series, and judges valid data based on the criterion of the mathematical model. Comprehensively judge the collected flame, temperature, humidity, gas and other information. Thereby eliminating a large amount of useless data and reducing the burden on the network. Table 1 shows the values of each item in the data fusion model.

Table 1

fire probability temperature (C) humidity gas flame image 3.023% 0 20% 30% none none 6.340% 10 60% 60% none none 19.980% 20 80% 80% none none 25.331% 30 100% 100% have have 55.326% 40 200% 300% have have

The present invention also builds a flame image recognition model and realizes an algorithm for recognizing flames. The author himself invented a low-power algorithm, which can recognize whether there is a flame in an image with very few hardware resources. After porting from the PC to the Arduino platform, it is found that the perfect combination of the characteristics of the platform has a significant effect on the improvement of node power consumption.

As shown in Figure 1-2, a multi-data fusion flame recognition device includes a sensing module, a communication module, an image acquisition module, and an arduino single-chip microcomputer, wherein the single-chip microcomputer includes a processing module, a data acquisition module, and a data fusion module. As shown in Figure 2, the data collected by various sensors is then obtained through data fusion.

Sensing module: collect basic data through temperature and humidity sensors, gas sensors, fire detection sensors and image sensors for data fusion;

Processing module: complete the data fusion calculation and image recognition algorithm through the Arduino single-chip computer, so as to judge the valid data and then pass it to the communication module. At the same time, the single-chip computer also has the function of power management to provide power strategy for the whole device;

Communication module: Xbee protocol stack implemented by hardware;

Data acquisition module: it is the driver program of the sensor, which drives the sensor to collect data, and automatically eliminates redundant, invalid, and mixed sensor data through the data gate value, so as to achieve the effect of power consumption;

Image acquisition module: used to acquire images, and then perform pattern recognition, which integrates the video driver and pattern recognition algorithms on Arduino;

Data Fusion Module: Combine all kinds of information into key information and report possible fires.

The temperature and humidity sensor is a DHT11 digital temperature and humidity sensor with calibrated digital signal output.

The gas sensor is based on the MQ3 gas sensor of the gas sensor, which is used to detect alcohol and ethanol gas in the air, and is used in combination with the Arduino special sensor expansion board to detect the concentration of smoke in the air.

The fire source detection sensor can detect light sources with wavelengths in the range of 760 nm to 1100 nm, convert them into digital signals and transmit them to the processing module.

The image sensor CMUcam4 is a programmable embedded image sensor, which uses Parallax P8X32A chip and Omini Vision9665CMOS camera module to capture photos of flames in the forest.

Referring to Figure 3, a multi-data fusion flame recognition method first collects data through a camera, then uses a single-chip microcomputer to delete valid data, deletes image information through gate values, and finally synthesizes valid graphic data to identify flames. Arduino and other single-chip microcomputers deployed on nodes have certain processing capabilities for data fusion. Taking into account the power and processing capabilities of nodes, a fusion algorithm based on coefficients is used. Assuming that there are multiple sensors, the output data of each sensor is X, and then use the weighted average method on them to obtain the fusion result.

Example 1

Simulated fire scene:

temperature humidity gas flame image 80 10% 200% - -

After receiving the temperature and humidity data, it is judged that there is a fire situation through the data fusion algorithm. The function of not turning on the flame and image recognition reduces the loss of energy efficiency. According to power estimation, the battery life of this device is more than five times that of ordinary nodes with the same function.

Example 2

Simulated fire scene:

temperature humidity gas flame image 0 10% 200% have -

After receiving the temperature and humidity data, it is impossible to judge whether there is a fire. The gas index is 200%. Through the fusion algorithm, it is concluded that the flame sensor needs to be turned on to identify the fire. After the device was opened, it was found that there was a fire, and the flame was successfully predicted.

Example 3

Simulated fire scene:

temperature humidity gas flame image 30 10% 20% have none

After receiving four items of data (temperature, humidity, gas, and flame), it is found that there may be a fire, and the fusion algorithm opens the last step to check whether there is a fire in the forest through the image recognition algorithm. The result is invalid data for the flame sensor. Arduino automatically judges not to enable wireless transmission, thus achieving the power saving effect.

The above is only the best implementation mode of the present invention, any simple changes or equivalent replacements of the technical solutions that can be clearly obtained by any person skilled in the art within the technical scope disclosed in the present invention all fall into the scope of the present invention within the scope of protection.

Claims (6)

1. A kind of flame identification device of multi-data fusion, is characterized in that, comprises sensing module, communication module, image acquisition module, arduino single-chip microcomputer, wherein single-chip microcomputer comprises processing module, data acquisition module, data fusion module: Sensing module: collect basic data through temperature and humidity sensors, gas sensors, fire detection sensors and image sensors for data fusion; Processing module: complete the data fusion calculation and image recognition algorithm through the Arduino single-chip computer, so as to judge the valid data and then pass it to the communication module. At the same time, the single-chip computer also has the function of power management to provide power strategy for the whole device; Communication module: Xbee protocol stack implemented by hardware; Data acquisition module: it is the driver program of the sensor, which drives the sensor to collect data, and automatically eliminates redundant, invalid, and mixed sensor data through the data gate value, so as to achieve the effect of power consumption; Image acquisition module: used to acquire images, and then perform pattern recognition, integrating the video driver and pattern recognition algorithms on Arduino; Data Fusion Module: Combine all kinds of information into key information and report possible fires. 2. The multi-data fusion flame identification device according to claim 1, wherein the temperature and humidity sensor is a DHT11 digital temperature and humidity sensor, which includes a calibrated digital signal output. 3. the flame recognition device of multi-data fusion according to claim 1, is characterized in that, described gas sensor is based on the MQ3 gas sensor of gas sensor, is used for detecting alcohol, ethanol gas in the air, and Arduino special-purpose sensor The expansion board is used in combination to detect the concentration of smoke in the air. 4. The multi-data fusion flame identification device according to claim 1, characterized in that, the fire source detection sensor can detect light sources with wavelengths in the range of 760 nm to 1100 nm, convert them into digital signals and transmit them to the processing module . 5. the flame recognition device of multi-data fusion according to claim 1, is characterized in that, described image sensor CMUcam4 is the embedded image sensor of programmable, adopts Parallax P8X32A chip and Omini Vision9665CMOS camera module to capture flame in forest Photo. 6. A flame identification method of multi-data fusion, characterized in that, first collect data through a camera, then use a single-chip microcomputer to delete and select effective data, delete and select image information by a gate value, and finally synthesize effective graphic data, thereby identifying flames.

Images