CN1139890C - Intelligent monitor system and method for grain insects in grain depot - Google Patents

Abstract

The present invention relates to an intelligent grain insect monitoring system and a method for monitoring grain insects in a grain depot. The system comprises a main computer in a monitor center, multi-channel front-end video monitoring stations at the front end of the grain depot, and image detection rods inserted into grains, wherein images of the grains and the grain insects are shot by the detection rods, the images are transmitted to the monitoring stations in the form of analog video signals and are digitalized, and the digitized video image data after digitalization is compressed; then the data is encoded for transmission and taken by the computer in the monitor center, and the data is identified and processed in different ways by the computer in the monitor center. The present invention is used for detecting insect generation situations during grain storage and management, and can be used as a research measure for observing and studying the growth conditions and growth principles of different grains and the grain insects in different varieties, and the technical problems are solved to realize the functions of automatic monitoring, automatic identification, automatic analysis, automatic numerical statement, automatic alarm, etc.

Description

Grain insects in grain depot intelligent monitor system and method

Technical field

The present invention relates to a kind of monitoring system and method, relate in particular to a kind of intelligence system and method that the grain worm of grain depot is monitored.

Background technology

State Council, State Planning Commission, State Grain Administration's decision in 1998 enlarge the grain reserves amount, intend in the whole nation newly-built 1,000 hundred million jin of grain depots.Existing about 1,200 hundred million jin of the national before this national grain depot and the local grain depot combined capacity of storing reach 2,200 hundred million jin after newly-built.But all these grain depots are not all realized grain worm automatic monitoring, the observation that grain is infested and find neither one effective means always.The grain depot managerial personnel have or not infestedly in order to survey in the grain, adopt grain worm trap method.The signal of grain worm trap is shown in detecting head (102) longitudinal diagram in Fig. 1 feeler synoptic diagram.Use metal tube or glass reinforced plastic pipe, tube wall is beaten the pore (104) that some grain worms can climb into, and the lower bottom part of pipe is equipped with chamfering plate washer (105), makes grain worm (106) crash into the back and is difficult to crawl out again, promptly becomes grain worm trap.The grain depot managerial personnel regularly or irregularly insert this trap pipe in grain, wait a period of time and think that the grain worm has entered, and take out the trap pipe again and watch or take back office and watch in detail.Traditional hand inspection method that Here it is.And, adopt the stifling way of regular poison routinely for deinsectization, do not have infested smoked termly yet, both increased the storehouse management cost, what all can have contamination to grain simultaneously, although being known as this poison can volatilize naturally, people's psychological effects always is difficult to floating.At this situation, native system is devoted to the grain worm is carried out automatic monitoring, and carries out Intelligent Recognition and analysis, in the hope of keeping promotes to some extent to grain depot foodstuffs.

Aspect the automation signal processing, existing a lot of proven technique, comprise digitizing, compressed encoding and communication code processing to video/signal of video signal, signal of video signal digitizing technique wherein, but reference book " Flame Image Process and analysis ", publishing house of Tsing-Hua University published in 1999, and Zhang Yujin writes; " multimedia system principle and application " People's Telecon Publishing House publishes nineteen ninety-five, and Hu Xiaofeng etc. write.Please note reference image input technology wherein.The image digitazation chip can use SAA7113 of Philips company or SAA7130 model.

Compressed encoding wherein can use the compression coding technology of MPEG-II standard, can be referring to international standard " MPEG-2 ISO/IEC 13818 ", published by ISO (International Standards Organization)/MPEG moving picture experts group assemble editing in 1994; The books typical case representative that the Chinese compiling is published is " new development of MPEG-2 moving picture experts group international standard and MPEG ", and publishing house of Tsing-Hua University publishes, and Zhong Yu chisels compilings such as professor.

Wherein computer network transmits, and uses conventional ethernet communication technology, can be referring to international standard IEEE802.

The Chinese book reference:

" computer network study course " Maritime Press published in 1991, Liang Zhenjun, Liang Bo compiling;

" Data Communication and Computer Network " Electronic Industry Press published in 1998, Yang Xinqiang, and Shao Junli writes;

" network communication software design guidelines " publishing house of Tsing-Hua University published in 1994, and Zhu Sanyuan etc. write.

The present invention uses multi-channel video monitoring station with above-mentioned function that video/signal of video signal is handled and the grain worm is carried out robotization detects and handle on the basis of these prior aries.

Summary of the invention

In order in the grain storage keeping, in time to find the grain worm, also for can close observation, study growing environment condition and the growth rhythm of different grains, variety classes grain worm, the invention provides a kind of grain insects in grain depot intelligent monitor system, not only can realize automatic monitoring to grain insects in grain depot, can also realize the grain worm is discerned automatically, analyzes automatically, adds up automatically, and automatic warning is provided.

For achieving the above object, grain insects in grain depot intelligent monitor system of the present invention is made up of the image probing rod of monitoring center's principal computer, silo front end multi-channel video monitoring station, insertion grain, wherein:

With network principal computer and one or more multi-channel videos monitoring station are coupled together, each multi-channel video monitoring station can connect a plurality of image probing rods according to the needs that robotization detects;

Each image probing rod inside comprises a camera, after the insertion grain, through being controlled to the picture illumination, take grain worm image, and the output analog video signal is to the multi-channel video monitoring station.

Video digitizer, compressed encoding, communication code processing are carried out to the vision signal of image probing rod output in the multi-channel video monitoring station, send into Surveillance center's principal computer through network again;

Described principal computer comprises:

Packed data that the receiver, video monitoring station is sent and the device of decoding and recovering;

To the background image that the video data reference that recovers is clapped when not having the grain worm by grain worm trap base plate in the image probing rod, the target image of taking when being monitored by the grain worm subtracts the device of getting to it;

To finding the solution the device that the method for threshold value is obtained threshold value through subtracting the image employing histogram of getting;

Calculate the device of the statistical value of images acquired and background image pixel colour difference;

Comparison means is used for summation and the difference of background image pixel chromatic value summation and the relation of threshold value of the pixel color value of comparison one frame images acquired;

Judgment means, when the difference of the summation of the pixel color value of a frame images acquired and background image pixel chromatic value summation during greater than threshold value, judging has the grain worm to occur in the image;

Warning device when judgment means has been judged the grain worm and occurred, is reported to the police immediately.

In the grain insects in grain depot intelligent monitor system of the present invention, described principal computer further comprises the device that separates grain worm physical form, and this device is when finding the grain worm, according to threshold value, with image binaryzation, separate grain worm and background image, extract the physical form of target grain worm.

In the grain insects in grain depot intelligent monitor system of the present invention, described principal computer also comprises the device that grain worm number is added up, and this device is ticked grain worm profile by adopting gradient method,, along searching for grain worm number is added up along silhouette edge.

In the grain insects in grain depot intelligent monitor system of the present invention, described principal computer also comprises data library device, store long-term observation data, grain worm activity video, be used for inquiry, statistics full storehouse year, season infested situation with the infested rule of different grains, and analyze humiture to the influence of grain worm growth and dosing deinsectization influence to the grain worm.

The present invention provides a kind of method that grain insects in grain depot is monitored simultaneously, may further comprise the steps:

Make image probing rod front end detecting head have grain worm trap function, imaging and control device are taken in installing in bar;

Image probing rod is inserted in the grain, after the grain worm climbs into, be controlled to the picture illumination and take grain worm activity video and microscopic morphology, and export analog video signal and carry out digitizing for the multi-channel video monitoring station, compression, and be sent to principal computer;

The packed data that principal computer receiver, video monitoring station is sent and the recovery of decoding;

With reference to the background image that grain worm trap base plate in the image probing rod is clapped when not having the grain worm, the target image of taking when being monitored by the grain worm subtracts it to be got;

To obtaining threshold value through subtracting the method that the image got adopts histogram to find the solution threshold value;

Calculate the statistical value of images acquired and background image pixel colour difference, have the grain worm to occur in the image if the difference of the pixel color value of a frame images acquired and background image pixel chromatic value greater than threshold value, is then judged, and report to the police immediately.The present invention has the following advantages:

1. survey automatically, reduce number of times and time that managerial personnel pass in and out silo, save the hand inspection workload;

2. but automatic monitor for continuously is regardless of and uninterruptedly monitors day and night, makes and finds the grain worm in good time immediately;

3. system can realize automatic identification, automatic counting, automatic warning, automatic statistical analysis;

4. system can store all monitoring situations, for regularly or irregularly study afterwards, provides grain keeping technology and the scientific research of grain worm to use.

Use native system can in time find grain worm in stock's grain, analyze insect pest situation, avoid or reduce the rotten loss that causes of grain; As do not have infestedly, and promptly blindly poison is stifling, and cost saving reduces the pollution to grain.

Description of drawings

Below in conjunction with drawings and Examples the present invention is elaborated.

Fig. 1 is the image probing rod synoptic diagram;

Fig. 2 is a grain insects in grain depot monitoring system pie graph;

Fig. 3 is the system monitoring main program flow chart;

Fig. 4 is data base querying master interface;

Fig. 5 is the system host master control interface;

Fig. 6 is the video decode software flow.

Embodiment1. system body is formed and principle of work

With reference to accompanying drawing 2, the native system main body is made up of three parts: monitoring center's principal computer (201), silo front end multi-channel video monitoring station (202), insert the image probing rod (203) of grain, all supporting in principal computer and the monitoring station have a corresponding software.

Host computer machine computer network couples together one or more monitoring stations (what of silo in the grain depot are number be decided by), and each monitoring station can connect a plurality of feelers according to the needs of automatic monitoring.Computer network is made up of thick coaxial cable (204), network transceivers (205), network terminal device (206) etc. among Fig. 2.By feeler picked-up grain and grain worm image, carry out digitizing through delivering to the monitoring station with the analog video signal mode, and to the video image data compression after the digitizing, be taken into by the center principal computer through communication code again, discern and all kinds of processing at background system.

System work process and principle of work are:

The grain depot staff inserts image probing rod (shown in Figure 1) in (or being fixed on) grain, for enable detection has or not in the grain infestedly, detecting head shell (102) is made grain worm trap, on the bar wall, stamp inclined hole (104), the bar inner chamber has air and attractant, the grain worm can actively become into.Device through the control lighting, can be taken grain worm image at the camera of feeler inside.The vision signal of camera output is sent into the video monitoring station, through video digitizer, compressed encoding, communication code, the machine network is sent into Surveillance center's principal computer as calculated again, handles, discerns, adds up, analyzes.

Below separately each several part specifies function and technology.2. overall system disposes and relevant index

2.1 for making a cover network system can manage a bed rearrangement grain depot, must adopt concentric cable to connect each monitoring station, thick cable root segment is 800～1000 meters, also can be prolonged by thick cable repeater in case of necessity.

2.2 a plurality of silos of next-door neighbour can be managed in a monitoring station, measuring point can be laid one or more in the storehouse, and measuring point generally is limited in 200 meters the wire length of monitoring station, if special requirement is arranged, can expand longlyer through special measure, but increases cost.

2.3 8 front end measuring points of a monitoring station general management during special requirement, can expand to 16, the total detection of total system is counted and is advisable to be no more than at 320.

2.4 automatically during itinerant monitor system, can carry out continuously or at random, also can regularly carry out, for example every time zero-time is 0 o'clock, 7 o'clock, 12 o'clock, 18 o'clock; To every sensing point continuous acquisition image in 10 second, 320 account for 3200 second time altogether.Make the automatic acquisition time of every day not take the work hours, do not influence artificial observation;

2.5 the frame per second of automatic image collection is no less than 20F/S, makes to observe the activity of grain worm continuously; The digitized resolution of every two field picture gets 640 * 480, reaches the precision of every millimeter about 20 pixels, can distinguish the thinnest grain worm completely; When micro-imaging, can reach the precision of every millimeter 400 pixel, make and to offer an explanation worm pin and worm back of the body decorative pattern;

2.6 for above-mentioned view data can effectively be transmitted under thick cable Ethernet 10Mbps bandwidth, Coding Compression Algorithm can be with video image data compression in＜4Mbps scope in the system video monitoring station.3. grain worm monitoring system operational scheme

With reference to accompanying drawing 3 system monitoring main program flow charts.

1. 2. 3. 4. 5. 6. the control reentry point of program circuit among the figure.

System is generally operational in the automatic monitoring mode, operation in 24 hours;

The needs that can adapt to managerial personnel, artificial participation system, semi-automatic monitoring, and query analysis afterwards adopt the man-machine interactively mode to operate.

3.1 patrol among the brake figure 301 automatically

Self-starting on time (3011), principal computer are according to system time, and startup entered this flow process automatically at 0 o'clock, 7 o'clock, 12 o'clock, 18 o'clock.

System starts 01 monitoring station automatically, and order opens 011,012 ... 018 feeler carries out in proper order to each feeler: open, the picked-up video image, mainly treatment step such as handle, cut out, return.Surveyed after last feeler of last monitoring station, returned 1. point of primary control program, waited next time arrival, perhaps entered personal monitoring or query analysis by artificial participation.

3012---picked-up video image implementation is: main frame is opened this feeler mains lighting supply by the video monitoring station, open the camera power supply, starting the video monitoring station makes by per second 20 frame rate pickup images and to video signal digitization, the packed data coding, the data communication coding, continue picked-up 10 seconds behind totally 200 two field pictures, camera power supply in the feeler is closed in the monitoring station, close mains lighting supply again, finish above-mentioned encoding process simultaneously, main frame is taken into whole 200 two field picture packed datas (5 key images frames wherein are set at least) by network from this monitoring station, recover source images through decompression.So far finish all videos image capturing process.

3013---main treatment scheme: main frame is stored the above-mentioned image that recovers temporarily, therefrom take out first key frame images, make background image (measuring when the system initialization) is subtracted and gets, the image that subtracts after getting is carried out the identification of grain worm, then calculate grain worm number if any the grain worm, be designated as x; As do not have the grain worm, then remember x=0; Then get the 2nd key frame images, repeat said process, have worm then to calculate borer population and put y, no worm is then put y=0; Get the 3rd key frame images again, repeat said process, have worm then to calculate borer population and put z, no worm is then put z=0.Relatively x, y, z three count, if not 0, then get the maximum and deposit database in, as the record of monitoring borer population, simultaneously above-mentioned video image is saved as file.Be the saving memory space, but the view data of store compressed.The image file name gives record in correspondence database.If x, y, z three numbers are all 0, then be designated as no worm in the database, simultaneously the compressing image data and the decompression image restored data of interim storage before the deletion.

If find to have worm,, intend picked-up grain worm micro-image in order to observe and analyze the kind of worm.Principal computer passes through motor in the selected video monitoring station control feeler herein, drives camera and presses close to object plane, and bust shot obtains grain worm micro-image, and exists in the image file, and filename is recorded in again in the correspondence database record.

3014---warning function

1. in case find that the grain worm grows out of nothing, warning immediately; Grain borer population amount is also pointed out warning than last time increasing;

2. computer screen is provided with zone of alarm, and flicker shows grain worm image, as alarm;

3. audible alarm continued to pipe 2 seconds every 1 minute, and the prompting people watch the infested situation of grain;

4. after warning is found, can manually close, withdraw from alarm condition.

3.2 personal monitoring's function (see in the accompanying drawing 3 302)

Clicking the master control menu with man-machine interaction mode enters.

Click corresponding monitoring station according to the want sensing point of monitoring and reach wherein sensing point, perhaps supervise the activity of grain worm, perhaps supervise grain worm form.

Check grain worm activity (3021)---this moment, main frame was lifted camera by motor in the selected monitoring station control feeler to get, can observe the object plane overall situation, the monitoring station does not stop to take out video image from camera with the 20F/S frame per second, do not stop compressed encoding, main frame does not stop to receive and decompress(ion) is play demonstration, and this promptly reaches real time monitoring grain worm activity usefulness.

Check grain worm form (3022)---be head, neck, leg feature and back texture, the color that Real Time Observation is found the grain worm, to analyze which kind of type of this Eimeria, at this moment must absorb the micro-image that amplifies, control camera for this reason and press close to the imaging object plane, specifically absorb grain worm close shot complete image.

3.3 query analysis function (see in the accompanying drawing 3 303)

This kind mode is to observe the situation of patrolling the survey record with statistical study in the past automatically by data and the image stored in the database afterwards.Wherein being divided into four kinds of working methods, all is to enter data base management system (DBMS) to carry out.

3031---check the former image file of depositing, with reference to accompanying drawing 4---data base querying grain worm monitoring situation interface, in window bottom input inquiry condition: year, month, day time and survey station number, survey period etc., if found to have the grain worm at that time, then correspondence can be pointed out video file name and image file name on this interface, click this document name, promptly demonstrate institute's video image of depositing and image.

3032---look into the borer population amount, utilize the interface of accompanying drawing 4, can find database Central Plains stock borer population quantitative changeization.

3033---look into the humiture situation, have the statistical value of measuring point environment temperature and humidity in this database equally, also can find by similar interface.Humiture and infested situation are put down in writing simultaneously, can assist the external condition of observing and analyzing the growth of grain worm.The measurement of temperature and humidity is not included in this paper scope.

3034---the insect pest situation statistical study, the long period accumulation by database can obtain the statistic analysis result about the monitoring of grain worm, and its function is:

1. statistical function: add up infested situation of full storehouse year

The annual infested situation in statistical storehouse

Adding up full storehouse divides season infested situation

Statistics native system monitoring situation

2. analytic function: can with grain storage condition, humiture situation, the Conjoint Analysis of dosing deinsectization situation

Analyze the growth of Various Seasonal grain worm, active situation

Analyze daytime, grain worm growth at night, active situation

Analyze different grain grain worm growths, active situation.4. database storing and management

Store in database patrolling the survey result automatically, be used for inquiring about and statistical study afterwards;

4.1 data item setting in the database:

Measuring point numbering---the unified numbering of all measuring points in the same system is comprising survey station numbering and measuring point sequence number.

Monitoring time: year, month, day, hour, min, get 2 for every;

Borer population

Report to the police: do not have/have

The storing image files name:

Storage source image file name:

Grains-type

The measuring point temperature

Measuring point humidity

Dosing deinsectization times last time 4.2 data base querying mode:

Press the measuring point inquiry

Press the monitoring time inquiry

By having/the alarm free inquiry

How much inquire about 5. system software explanations by borer population

System cooperates hardware to realize that the software of above-mentioned functions has following a plurality of module:

System master interface and main control software;

Grain worm statistics, alarm software;

Grain worm target is extracted and Counting software;

Grain worm image video encoding software;

Grain worm image video decode software;

System data library inquiry and statistical analysis software.

5.1 system master interface and main control software

System master interface as shown in Figure 5, left side main part is that image shows main window among the figure, demonstrates the existing moth that adds red circle mark in the grain among the figure.The right is 6 feature operation buttons from top to bottom among the figure.Function is respectively:

501 select the monitoring station, by demonstrating later all video monitoring stations numberings that system connects, click arbitrary numbering, promptly enter the monitoring range of this station administration;

502 select the monitoring point, image probing rod of the corresponding use in each monitoring point; Click wherein arbitrary numbering, promptly begin to control to obtain and select the grain worm image that shooting comes from this;

503 data base queryings, data base querying window as shown in Figure 4 appears after the click, the Monitoring Data that many time point monitorings are obtained before checking therein, comprising: the grain borer population order that monitoring time, monitoring station number, monitoring period, automatic analyser number go out also comprises the video file name, the static picture document name that write down the image of clapping.

504 check grain worm number, occur the data base querying window after the click and make the current grain worm number of the corresponding sensing point of direct demonstration;

The cereal temperature that the data base querying window directly shows current website appears in 505 measuring point grain temperature after the click;

The grain moisture that the data base querying window directly shows current website appears in 506 measuring point humidity after the click;

The master control interface of Fig. 5 is manipulated for man-machine interactively is provided, and provides two kinds of modes of operation with primary control program flow process shown in the accompanying drawing 3.

5.2 grain worm statistics alarm software

Touring the measuring of each image probing rod that monitoring center's principal computer is connected each video monitoring station automatically analyzed the image of getting into, adopts the statistical mathematics method during analysis.Calculate images acquired and the statistical value that is compared the image pixel colour difference.There is the grain worm to occur in the image if statistical value greater than threshold values, is judged, reports to the police immediately.Type of alarm has two kinds: a kind of is to utilize computer screen to report to the police, and screen is provided with zone of alarm, and flicker therein shows grain worm image, and shows the numbering of this monitoring station and monitoring point.Another kind is an audible alarm, continues to pipe 2 seconds every 1 minute.After warning is found, can manually close warning, withdraw from alarm condition.

5.3 grain worm target is extracted and Counting software

1. background subtraction

Grain worm trap base plate in the image probing rod is clapped into background image when the system initialization, and the target image of taking when being monitored by the grain worm reduces background image.Below be that the image after the subduction is handled.

2. target is extracted

It is target grain worm to be extracted through the image of subduction from above-mentioned that target is extracted.The first step is asked for threshold values according to the histogram of image.Method is:

Because processed image mainly is two types of target and backgrounds, so obtain threshold values according to the expectation variance of image, effect is more satisfactory, and promptly the probability that occurs according to each gray-scale value determines that formula is:

Thresh=∑ gray*p[gray]; Wherein gray is [0..255], p[gray] for the image intermediate value probability of the gray-scale value appearance of gray.

Second step is according to threshold values, with image binaryzation.Be gray-scale value less than threshold values, putting its gray-scale value is 0, greater than threshold values, putting its gray-scale value is 255.So just grain worm and background separation can have been extracted the physical form of target grain worm.

3. Target Recognition and counting

Target Recognition is meant the grain worm in the image is split, and earlier target is extracted outer contour, and outline line is sought counting by certain rule.Concrete grammar is: begin search from first pixel of image, if search the point (being that pixel value is 0 point) on the profile, then should put stacked, check this some pixel point value of eight directions on every side then successively, see that whether they be point on the profile, if, then again that it is stacked, otherwise, not stacked.Simultaneously the isolated point that looses is rejected.Only return and find the initial profile point, illustrate that outline line is complete, storehouse also reaches a great deal of, can count one or two grain worm.All outline lines of one two field picture all travel through and finish, and then grain worm counting is finished.

5.4 video decode software

The decoding implementation procedure as shown in Figure 6.Wherein

601 variable-length decodings

Variable-length decoding solves harbour information and DCT data by look-up table.

602 reverse scannings

The reverse scanning process is with one-dimensional data QFS[n] be transformed to two-dimentional coefficient arrays QFS[v] [u], wherein the scope of n is 0-63, the scope of u, v is 0-7.

603 re-quantization processes, wherein each link be respectively re-quantization arithmetical operation, amplitude limit, the control that do not match, division is as follows:

6031 re-quantization arithmetical operations:

The inverse quantization method of Intra-coded blocks DC coefficient is different with the quantization method of other coefficient.In frame in the piece, F " [0] [0] should be by QF[0] [0] constant intra_dc_mult of multiply by by the intra_dc_precision decision obtains.Other coefficient is rebuild F by the arithmetical operation of following equation definition " [v] [u].

F”[v][u]＝((2QF[v][u]+k)×W[w][v][u]×quantiser_scale)/32

Wherein K is 0 during piece in frame, is sign (QF[v] [u]) during piece in the non-frame.Weighting matrix W[w] w value among [v] [u] is 0-3, indicates current matrix by color-difference formats and block type decision.The suitable quantiser_scale of q_scale_type decision in Quantiser_scale_code and the picture coding expansion.

6032 amplitude limits and the control of non-coupling

The coefficient that amplitude limit is obtained by the re-quantization arithmetical operation is so that it is in scope [2048:+2047].Therefore:

In to piece all rebuild, after coefficient F ' [v] [u] summation of amplitude limit, and if be even number, then tackle coefficient F[7] [7] make amendment.If F ' [7] [7] is an odd number, F[7] [7] equal F ' [7] [7] and subtract one; If F ' [7] [7] is an even number, F[7] [7] equal F ' [7] [7] and add one.This non-coupling control is that IDCT detunes control.

604 inverse DCTs, calculate as follows: $f(i,j)=\frac{1}{4}{\mathrm{\Σ}}_{u=1}^8{\mathrm{\Σ}}_{v=1}^{8}c(u,v)\·F(u,v)\cos \frac{2i+1}{16}\mathrm{u\π}\cos \frac{2j+1}{16}\mathrm{v\π},$ I wherein, j=0,1 ... 7, u, v=0,1 ... 7.

605 motion compensation

The key of motion compensation is how to utilize motion vector to rebuild original image.In decoder end, rebuild the difference motion vector by the parameter in the code stream.For decoding motion vectors, demoder keeps 4 motion vector predictor (each of horizontal component and vertical component), is expressed as PMV[r] [s] [t].To each prediction, at first obtain vector ' [r] [s] [t].To each chrominance component, provide motion vector vector[r] [s] [t], carry out convergent-divergent according to color-difference formats then.6. system extension function

According to safety monitoring and grain depot management work needs, can on native system front end video monitoring station, increase and connect the logical video camera of 1 Daepori, place outside the silo grain video camera suitably highly local, can monitor that the grain heap is high in the storehouse, calculate stored number, can monitor simultaneously to enter personnel and active situation in the storehouse.

Claims (10)

1. grain insects in grain depot intelligent monitor system is characterized in that it is made up of monitoring center's principal computer, silo front end multi-channel video monitoring station, the image probing rod that inserts grain, wherein:

With network principal computer and several multi-channel video monitoring stations are coupled together, each multi-channel video monitoring station can connect a plurality of image probing rods according to the needs that robotization detects;

Utilize image probing rod to take grain worm image in the grain, and the output analog video signal is given the multi-channel video monitoring station;

The multi-channel video monitoring station is collected the vision signal of image probing rod output and is carried out video digitizer, compressed encoding, communication code processing, sends into Surveillance center's principal computer through network again;

Described principal computer comprises:

Packed data that the receiver, video monitoring station is sent and the device of decoding and recovering;

To the background image that the video data reference that recovers is clapped when not having the grain worm by grain worm trap base plate in the image probing rod, the target image of taking when being monitored by the grain worm subtracts the device of getting to it;

To finding the solution the device that the method for threshold value is obtained threshold value through subtracting the image employing histogram of getting;

Calculate the device of the statistical value of images acquired and background image pixel colour difference;

Comparison means is used for summation and the difference of background image pixel chromatic value summation and the relation of threshold value of the pixel color value of comparison one frame images acquired;

Judgment means, when the difference of the summation of the pixel color value of a frame images acquired and background image pixel chromatic value summation during greater than threshold value, judging has the grain worm to occur in the image;

Warning device when judgment means has been judged the grain worm and occurred, is reported to the police immediately.

2. grain insects in grain depot intelligent monitor system as claimed in claim 1 is characterized in that: described image probing rod is equipped with inclined hole on its bar wall, the bar inner chamber is equipped with air and attractant, make the grain malnutrition due to parasitic infestation utmost point become into.

3. grain insects in grain depot intelligent monitor system as claimed in claim 2 wherein uses image probing rod to insert in the grain, after the grain worm climbs into, enables to take two kinds of images, and a kind of is that minute surface is far away from object plane, takes grain worm active situation; Another kind is minute surface and object plane close together, obtains local grain worm form.

4. grain insects in grain depot intelligent monitor system as claimed in claim 1 wherein uses a cover to connect network mode for 2 grades, and the first order connects each monitoring station by principal computer through concentric cable, makes overall networking scope bigger; The second level is by each feeler of monitoring station Y-connection.

5. grain insects in grain depot intelligent monitor system as claimed in claim 1, it is characterized in that: described warning device utilizes computer screen to report to the police and the audible alarm dual mode, screen is provided with zone of alarm, the display alarm image that glimmers therein, and show the numbering of corresponding monitoring station and monitoring point; The audible alarm mode then is to continue to pipe 2 seconds every 1 minute.

6. grain insects in grain depot intelligent monitor system according to claim 1, it is characterized in that: when using the multi-channel video monitoring station that grain worm image is carried out digitizing, video frame rate is no less than 20F/S, and image resolution-ratio reaches 640 * 480, be transmitting data in real time, high data compression ratio 〉=200.

7. grain insects in grain depot intelligent monitor system as claimed in claim 1, wherein said principal computer also comprise the device that separates grain worm physical form, and this device is when finding the grain worm, according to threshold value, with image binaryzation, separate grain worm and background image, extract the physical form of target grain worm.

8. grain insects in grain depot intelligent monitor system as claimed in claim 1, wherein said principal computer also comprises the device that grain worm number is added up, this device is ticked grain worm profile by adopting gradient method,, along searching for grain worm number is added up along silhouette edge.

9. grain insects in grain depot intelligent monitor system as claimed in claim 1, wherein said principal computer also comprises data library device, store long-term observation data, grain worm activity video, be used for inquiry, statistics full storehouse year, season infested situation with the infested rule of different grains, and analyze humiture to the influence of grain worm growth and dosing deinsectization influence to the grain worm.

10. method that grain insects in grain depot is monitored may further comprise the steps:

Make image probing rod front end detecting head have grain worm trap function, imaging and control device are taken in installing in bar;

Image probing rod is inserted in the grain, after the grain worm climbs into, be controlled to the picture illumination and take grain worm activity video and microscopic morphology, and export analog video signal and carry out digitizing for the multi-channel video monitoring station, compression, and be sent to principal computer;

The packed data that principal computer receiver, video monitoring station is sent and the recovery of decoding;

With reference to the background image that grain worm trap base plate in the image probing rod is clapped when not having the grain worm, the target image of taking when being monitored by the grain worm subtracts it to be got;

To obtaining threshold value through subtracting the method that the image got adopts histogram to find the solution threshold value;

Calculate the statistical value of images acquired and background image pixel colour difference, have the grain worm to occur in the image if the difference of the pixel color value of a frame images acquired and background image pixel chromatic value greater than threshold value, is then judged, and report to the police immediately.

Images