CN101900687B - Method for monitoring and early warning water bloom in small water area based on image processing - Google Patents

Abstract

The invention provides a method for monitoring and early warning water bloom in a small water area based on image processing, which relates to a method for monitoring and early warning water bloom in a small water area. In the invention, a camera and a computer are used for realizing the purposes of monitoring and early warning the water bloom in the small water area through calculation according to programs. The invention has the characteristics of simple monitoring equipment, strong environment adaptive capacity, small algorithmic calculated amount, high speed, low requirements for hardware, flexibility, convenience, low cost, convenient maintenance and the like, can be directly embedded with the existing water surface video monitoring system or can be developed into embedded equipment or can directly operate on the computer, and can realize the purpose of quickly and effectively early warning the water bloom in real time. The invention can be widely used for monitoring and early warning the water bloom in small water areas such as upstream reservoirs of rivers, reservoirs used as sources of drinking water of towns, river reaches in water taking areas of towns, rivers, lakes or landscape water areas where the water bloom often occurs, and the like.

Description

A kind of method for monitoring and early warning water bloom in small water area based on Flame Image Process

Technical field

The invention belongs to the monitoring water environment technical field, be specifically related to the wawter bloom monitoring and pre-alarming method of small-sized waters (be the reservoir, section, water district, cities and towns of river, river upper pond, drinking water source, cities and towns, often river, river, lake country and the view waters etc. of wawter bloom take place).

Background technology

In recent years; Along with industrial, scientific and technological develop rapidly; The a large amount of worker of containing a large amount of nitrogen, phosphorus, agricultural and domestic refuses enter in the water; Cause the fresh water in waters such as river, river, lake eutrophication to occur and break out wawter bloom (like green alga wawter bloom, blue-green alga bloom etc.); Wawter bloom all took place in the Taihu Lake of China, Dian Chi, Chaohu, Hongchehu Lake, and it is that maximum harm is that wawter bloom causes: 1. algae toxin is through the carcinogen direct threats human beings'health and the existence of the generation of food chain remote effect human beings'health or algae, polluted source; 2. the filtration unit of waterworks is by algae " wawter bloom " filling, and influence is fetched water; 3. " wawter bloom " that swim on the water surface influences view, and unpleasant stink, contaminated environment etc. are arranged.It is thus clear that the wawter bloom phenomenon frequently occurs, not only cause the destruction of large tracts of land water environment, and brought enormous economic loss.Therefore; In the river, the waters, source at the upper reaches, river, local lake and fresh water sources such as (water district, cities and towns, drinking water source and view waters etc.), specific function waters monitor, it is very significant that generation, variation and the condition of a disaster trend of this waters wawter bloom are carried out monitoring and warning.

Existing wawter bloom monitoring and pre-alarming method; Disclosed like on August 6th, 2008; Publication number is CN101236519A's " buoy that is used for blue algae monitoring and blue algae bloom prealarming " patent, and disclosed buoy is made up of buoy carrier, instrument room, monitoring sensor aggregate (being formed by five kinds of different sensor set), communication antenna, solar panel.During application, the buoy of some is arranged in the ad-hoc location in tested waters, the different depth of water body is monitored, and with data monitored through antenna transmission to monitoring center, carry out real-time early warning after the data analysis.The major defect of this patent is:

1. the limitation of monitoring is big.This patent only is arranged in the buoy of some the ad-hoc location in tested waters; Different depth in the water body is monitored (being underwater monitoring); Can not monitor the larger area waters, can not monitor the Diffusion Law of waterborne contaminant, distribution range, pollution level etc.; Buoy need directly contact with water body in monitoring, and occupies certain water space, and directly use in the waters (like drinking water source, view waters etc.) that is not suitable in some specific functions.Therefore, the limitation of buoy monitoring is bigger.

2. maintenance cost is high, and the monitoring cost is high.This patent needs special technician regularly to monitoring is on-the-spot buoy to be safeguarded in practical application, consumption manpower consuming time, thus increased maintenance cost significantly.In order to obtain water quality information accurately, just must use expensive high precision monitor sensor, and the sensing of this patent is broad-minded, kind is many, thereby has improved the monitoring cost.

3. popularization is poor.Because the cost of this patent is high, complex structure, use, operation and maintenance highly professional strong, be unfavorable for popularizing and use in administrative authority, enterprise and R&D institution.

Summary of the invention

The objective of the invention is deficiency to existing wawter bloom monitoring and pre-alarming method; A kind of method for monitoring and early warning water bloom in small water area based on Flame Image Process is provided; It is low to have cost; Characteristics such as it is strong that environment is suitable for ability, can obtain marine pollution information in real time and can discern wawter bloom automatically and report to the police, easy to utilize.

Mechanism of the present invention: the present invention is that sample figure sets up the color prior model with the historical wawter bloom area image in tested waters at first; Each pixel that is about to sample figure from three primary colors (RGB) space conversion to look, show the score and leave (HSV) space; In order under various light conditions, to lock the wawter bloom zone; Therefore abandon saturation degree (S) and brightness (V) component in the HSV color space, only adopt tone (H) component.The HSV model is to press close to human perceptive mode to color most; And colouring information is relatively more responsive to H (tone) component in this space; And H (tone) component has been removed the influence of illumination largely, the influence that the color prior model of therefore setting up is not changed by ambient light.

Set up the color prior model in wawter bloom zone earlier, establish x ' _j} _J=1,2...nFor n the pixel in the wawter bloom of historical sample figure zone, in color model, when the discretize value of H (tone) component is that the tone probability of i (i=1,2..., 360) is:

p＝{p _i} _{i＝1，2…m}； $p_i=\frac{\underset{j=1}{\overset{n}{\mathrm{\Σ}}}{\left|\right|x_j^{\′}\left|\right|}^2\mathrm{\δ}[b\left(x_j^{\′}\right)-i]}{\underset{j=1}{\overset{n}{\mathrm{\Σ}}}\left({\left|\right|x_j^{\′}\left|\right|}^2\right)};$ $\underset{i=1}{\overset{m}{\mathrm{\Σ}}}{p}_i=1---\left(1\right)$

In the formula (1): p is a color model, p _iBe that tone value is the probability of i, δ is the δ function, function b (x ' _j) be space R ²→ { 1,2 ... The index of m} promptly is positioned at position x ' _jPixel to the index in histogram quantization characteristic space.

Use continuous adaptive property average drifting (Camshift) algorithm to detect the wawter bloom color lump again; Continuous adaptive property average drifting algorithm is to be developed by average drifting (MeanShift) algorithm, and the process of average drifting (MeanShift) algorithm is exactly through the sampling mean shift vector M on the nuclear G _{H, G (x)}(promptly examining the estimation of the density gradient on the K) upgraded a recursive procedure at nuclear G center.The average drifting algorithm application begins computation of mean values translation vector constantly by the initial center of position of nuclear G when continuous sequence, iteration is new target location more, up to converging on the Optimum Matching point.

M _{H, G (x)}(promptly examining the estimation of the density gradient on the K) computing formula is following:

$M_{h,G\left(x\right)}=\frac{h^2\▿{\hat{f}}_K\left(x\right)}{2/C{\hat{f}}_G\left(x\right)}---\left(2\right)$

Wherein: h is a bandwidth, and C is a normaliztion constant, and the sampling mean shift vector on the visible nuclear G is the estimation of the density gradient on the nuclear K. is the multivariate density Estimation of h for nuclear K goes up bandwidth:

${\hat{f}}_K\left(x\right)=\frac{1}{n{h}^d}\underset{j=1}{\overset{n}{\mathrm{\Σ}}}k\left({\left|\right|\frac{x-x_i}{h}\left|\right|}^2\right)---\left(3\right)$

Set x ' _j} _{J=1,2 ... N}Be d dimension Euclidean space R ^dN point, k (x) representes the kernel function of this pixel:

Wherein: c _dBe d dimension unit spheroid volume.

Continuous adaptive property average drifting algorithm is exactly travel direction projection and average drifting calculating in the processing region of video image; With the barycenter of present frame search window and area initial value, carry out iteration again and the detection of target is followed the tracks of realizing as the next frame search window.

Filter with improved continuous adaptive property average drifting algorithm then and disturb color lump; Promptly before using the Camshift algorithm, at first probability distribution graph is carried out the burn into expansion process, remove the noise piece; Detect the inside and outside contour of target color lump then, find the cavity in the target color lump to fill.Yet after carrying out above pre-service, possibly still have the interference color lump, so be applied to be necessary to improve the adaptability of Camshift algorithm when wawter bloom detects.

Algae occurs from the water surface and gather wawter bloom and break out comprehensively, need about 15 days time usually, so within a short period of time; Occurring the green area variation on the water surface is slowly; Comparatively speaking, the adjacent two two field picture time intervals are very short (a few tens of milliseconds), and the target color lump is very little in the area change of interframe so; Based on this Changing Pattern, we just can come overanxious interference color lump through the predicted characteristics Method for Area.

If the area of color lump is S in the current frame image _n, the prediction area of next frame is S ' _N+1, so

S′ _n+1＝S _n+A (5)

Wherein, constant A is an empirical value (being provided with as the case may be), the increment of expression interframe target color lump area.Just can predict the area of target color lump in the next frame image according to formula (5), if satisfy

|S _n+1-S′ _n+1|≤ξ (6)

S _N+1The actual measurement area of expression next frame color lump, the error range of ξ for allowing, explain current detection to color lump meet the Changing Pattern of interframe color lump area; If | S _N+1-S ' _N+1| when exceeding in the error range of permission, then think the next frame actual detected to the color lump area do not meet the Changing Pattern of interframe wawter bloom region area, redundance is defined as disturbs color lump and it is filtered.

The technical scheme that realizes the object of the invention is: a kind of method for monitoring and early warning water bloom in small water area based on Flame Image Process; Utilize camera and computing machine,, set up the priori color model according to historical data through program; Use improved continuous adaptive property average drifting (Camshift) algorithm to filter out and disturb color lump and detect the wawter bloom zone; Through the mutation analysis the condition of a disaster trend of this region area, and set up the Early-warning Model that is similar to meteorological disaster, monitor and the early warning wawter bloom.Concrete steps are following:

(1) sets up the color prior model

At first, set up the color prior model, that is: according to the historical wawter bloom in tested waters regional sample figure or video data

If x ' _j} _{J=1,2 ... N}Be n pixel of sample figure or video data, with each pixel from RGB (three primary colors) space conversion to HSV (look, show the score from) space, be i (i=1 with the discretize value of H (tone) component; 2...; 360), zoom to [0 to the scope [0,360] of H (tone) component; 255], so as the value of span can represent with a byte (byte).Be calculated as follows the tone probability of H component through computing machine:

p＝{p _i} _i＝12…m； $p_i=\frac{\underset{j=1}{\overset{n}{\mathrm{\Σ}}}{\left|\right|x_j^{\′}\left|\right|}^2\mathrm{\δ}[b\left(x_j^{\′}\right)-i]}{\underset{j=1}{\overset{n}{\mathrm{\Σ}}}\left({\left|\right|x_j^{\′}\left|\right|}^2\right)};$ $\underset{i=1}{\overset{m}{\mathrm{\Σ}}}{p}_i=1---\left(1\right)$

Wherein: p is a color model, p _iBe that tone value is the probability of i, δ is the δ function, function b (x ' _j) be space R ²→ { 1,2 ... The index of m} promptly is positioned at position x ' _jPixel to the index in histogram quantization characteristic space.

(2) the wawter bloom zone is detected

1. small-sized waters is monitored

After the completion of (1) step; Camera is set up on the bank of (being the upper pond in river, river or the reservoir of drinking water source, cities and towns or the section of water district, cities and towns or river, river, lake waters or the view waters etc. that wawter bloom often takes place) in small-sized waters, and is connected with computing machine through video transmission line.The particular location that camera sets up, quantity and height are confirmed according to the concrete condition in tested waters.In order to the water surface situation of monitoring tested waters (changing the most significantly because of the breakout of water bloom in waters is the variation of water surface visual signature) and absorb the video image of the water surface, reach the purpose of real-time monitoring Surface Picture.

2. back projection is handled

After the completion of (2)-1. step; Each frame video image to the input of (2)-1. step carries out back projection processing (Back Projection); Promptly to each pixel in the video image processing region, the matching degree of the color prior model of setting up through this pixel of computer inquery and (1) step (being H component tone probability model), (other the regional probability outside this zone are 0 just to obtain the probability that this pixel is an object pixel; The value of each pixel has just become a kind of discretize tolerance that color of object information appears at the possibility here in the image; The possibility that here occurs is big more, and the value of pixel is just big more, otherwise then more little).Through above-mentioned processing, just obtain the color of object back projection figure of every two field picture.

3. detect and cut apart the wawter bloom color lump and filter the interference color lump

(2)-2. step was detected the wawter bloom color lump and disturbs color lump to handle with filtering, that is: after accomplishing

The back projection figure that earlier (2)-2. step was obtained carries out burn into and expands, and detects the inside and outside contour of target color lump and finds cavity in the target color lump to fill and wait pre-service, so just can remove a part of interference color lump, avoids noise to a certain extent.

Detect again and cut apart the wawter bloom color block areas, promptly press following formula constantly to the back projection figure computation of mean values translation vector M after the above-mentioned pre-service of process through computing machine _{H, G}(x):

$M_{h,G\left(x\right)}=\frac{h^2\▿{\hat{f}}_K\left(x\right)}{2/C{\hat{f}}_G\left(x\right)}---\left(2\right)$

Wherein: h is a bandwidth; C is a normaliztion constant,

be that the last bandwidth of nuclear K is the multivariate density Estimation of h:

${\hat{f}}_K\left(x\right)=\frac{1}{n{h}^d}\underset{j=1}{\overset{n}{\mathrm{\Σ}}}k\left({\left|\right|\frac{x-x_i}{h}\left|\right|}^2\right)---\left(3\right)$

Set x ' _j} _{J=1,2 ... N}Be d dimension Euclidean space R ^dN point, k (x) representes the kernel function of this pixel:

Wherein: c _dBe d dimension unit spheroid volume.In order to the position of iteration renewal target color lump, up to the Optimum Matching point that converges on, the zone of this match point is exactly the target color lump (being the wawter bloom color block areas) in the current frame image.

Judge: when not detecting the target color lump, think that then wawter bloom does not appear in the water surface, returned for (2)-1. step and obtain next frame and proceed to detect; When detecting the target color lump, then with the barycenter of present frame Optimum Matching point and area as the initial value of next frame search window, proceed iterative computation and find the Optimum Matching point.So till loop iteration to the EOP (end of program), just detect the target color lump that is partitioned into every two field picture.

Filter then and disturb color lump; The area of the present frame that the last step of i.e. calculating obtains and the target color lump of next frame; Interframe Changing Pattern according to the wawter bloom region area; Utilize the target color lump area of present frame to predict the target color lump area in the next frame image, promptly calculate the difference (being redundance) between next frame color lump actual detected area and the next frame prediction area, judge according to following formula:

|S _n+1-S′ _n+1|≤ξ (5)

Wherein: S _N+1Be the actual measurement area of next frame color lump, S ' _N+1Be the prediction area of next frame, the error range of ξ for allowing.

When | S _N+1-S ' _N+1| during≤ξ, then think the next frame actual detected to the color lump area meet the interframe Changing Pattern of wawter bloom region area and with it as testing result; When | S _N+1-S ' _N+1|＞ξ, when promptly difference exceeds in the error range of permission, then think the next frame actual detected to the color lump area do not meet the Changing Pattern of interframe wawter bloom region area, redundance is defined as disturbs color lump and it is filtered.

Through after the above-mentioned processing, just detect to be partitioned into and filter the wawter bloom color lump that disturbs behind the color lump, obtain the elemental area of wawter bloom color lump in image.Then testing result is demarcated and exported to the wawter bloom color block areas in every two field picture.

(3) bloom prealarming

1. calculate wawter bloom zone real area:

(2) step calculated (2) earlier and goes on foot the number percent that the elemental area of wawter bloom color lump in image that obtains accounts for the visual field pixel total area after accomplishing, and the actual total area that multiply by the visual field just can obtain wawter bloom zone real area.Pass through camera calibration again; Calculate the actual total area in the visual field; The object of reference that is about to known area places the position, waters under the camera, calculates the ratio that its elemental area in image accounts for the visual field pixel total area, just obtains the actual total area in the visual field divided by this ratio with the real area of this known object of reference; And the wawter bloom that obtains zone real area outputed on the computer software interface, so that the user gets information about the real area in wawter bloom zone.

2. set up Early-warning Model:

After the completion of (3)-1. step; Go on foot the wawter bloom zone real area that obtains according to (3)-1.; Be at interval by 6～10 hours earlier, calculate the average of the wawter bloom zone real area in each time interval, set up the one-variable linear regression forecast model through the Change in Mean rule of this real area of Computer Analysis again; Predict the average of the real area in the next time interval; And the average of calculating this real area accounts for the number percent of the waters total area, and the wawter bloom disaster degree with this number percent representative is a foundation then, and the advanced warning grade of breakout of water bloom is divided into blueness, yellow, orange and red four grades; Corresponding early warning interval be respectively generally, heavier, serious and four grades of especially severe; And judge:, then turn back to and returned for (2)-1. step and obtain next frame and proceed detection when the number percent of the wawter bloom disaster degree of representative during less than early warning value (size of early warning value confirms that according to the concrete condition in tested waters and application demand the present invention gets 5%); When the number percent of wawter bloom disaster degree of representative during, then carry out early warning according to the advanced warning grade of dividing greater than early warning value.The concrete division of breakout of water bloom advanced warning grade is following:

General blue early warning (being fragmentary property wawter bloom): wawter bloom is sporadicly gathered, and main waters district algae bio density is less than 3,000 ten thousand/L, and the wawter bloom area is more than or equal to 5% of the water body total area.

Heavier yellow early warning (being the locality wawter bloom): algae gathers in local waters, and main waters district algae bio density is between 3000～5,000 ten thousand/L, and the wawter bloom area is more than or equal to 10% of the water body total area.

Serious orange early warning (being regional wawter bloom): when generation area property wawter bloom, main waters district algae bio density is between 5000～8,000 ten thousand/L, and the wawter bloom area is more than or equal to 40% of the water body total area.

Especially severe red early warning (being comprehensive wawter bloom): wawter bloom is comprehensive to be broken out, and the algae bio density in main waters district is greater than 8,000 ten thousand/L, and the wawter bloom area is more than or equal to 60% of the water body total area.

After the present invention adopts technique scheme, mainly contain following effect:

1. has good popularization.The invention provides a kind of monitoring and pre-alarming method of wawter bloom fast and effectively; Can discern the wawter bloom zone automatically and calculate this regional real area in real time, can carry out early warning so that notice relevant departments understand wawter bloom disaster degree in real time and take corresponding control measures according to advanced warning grade.In practical application, directly implant in traditional water surface video frequency monitoring system or be developed to embedded device and can both realize monitoring and warning wawter bloom, have intellectuality, characteristics such as easy to operate, be convenient to popularize and use in administrative authority, enterprise and R&D institution.

2. cost is low.The algorithm computation amount that the present invention adopts is little, speed is fast, and is lower to hardware requirement.In the practical application, no matter directly operation on computers still is developed to embedded device, can effectively realize the wawter bloom monitoring and warning real-time, has practiced thrift the monitoring cost to a great extent, and is flexible.

3. strong to adaptive capacity to environment.The present invention utilizes the camera monitoring water surface and removes the influence of illumination to a great extent; The noise that complicacy and the ambient light variation that can effectively avoid aquatic environment etc. brings; Thereby can conform variation well, so monitoring equipment is simple, cost is low; Be convenient to safeguard, and can monitor effectively in real time.

The present invention can directly implant in traditional water surface video frequency monitoring system or be developed to embedded device, is widely used in reservoir, the section, water district, cities and towns of river, river upper pond, drinking water source, cities and towns or the wawter bloom monitoring and warning in small-sized waters in river, river, lake country and the view waters etc. of wawter bloom often takes place.

Description of drawings

Fig. 1 is the program flow chart of the inventive method.

Embodiment

Below in conjunction with embodiment, further specify the present invention.

Embodiment

As shown in Figure 1; A kind of method for monitoring and early warning water bloom in small water area based on Flame Image Process; Reservoir to Peng Xi river, Kai Xian, Chongqing City (being positioned at the one-level tributary, the Changjiang river in area, vital organs, reservoir area of Three Gorges) carries out the wawter bloom monitoring and warning, about 6.3 ten thousand square metres of the total area in this section waters, and concrete steps are following:

(1) sets up the color prior model

At first, set up the color prior model, that is: according to the historical wawter bloom in tested waters regional sample figure or video data

If x ' _j} _{J=1,2 ... N}Be n pixel of sample figure or video data, with each pixel from RGB (three primary colors) space conversion to HSV (look, show the score from) space, be i (i=1 with the discretize value of H (tone) component; 2...; 360), zoom to [0 to the scope [0,360] of H (tone) component; 255], so as the value of span can represent with a byte (byte).Be calculated as follows the tone probability of H component through computing machine:

p＝{p _i} _{i＝1，2…m}； $p_i=\frac{\underset{j=1}{\overset{n}{\mathrm{\Σ}}}{\left|\right|x_j^{\′}\left|\right|}^2\mathrm{\δ}[b\left(x_j^{\′}\right)-i]}{\underset{j=1}{\overset{n}{\mathrm{\Σ}}}\left({\left|\right|x_j^{\′}\left|\right|}^2\right)};$ $\underset{i=1}{\overset{m}{\mathrm{\Σ}}}{p}_i=1---\left(1\right)$

Wherein: p is a color model, p _iBe that tone value is the probability of i, δ is the δ function, function b (x ' _j) be space R ²→ { 1,2 ... The index of m} promptly is positioned at position x ' _jPixel to the index in histogram quantization characteristic space.

(2) the wawter bloom zone is detected

1. small-sized waters is monitored

After (1) step accomplished, on Kai Xian, Chongqing City splashes the water tower on bank, small-sized waters of small stream River Reservoir, set up a camera, the camera visual field covers whole waters, and is connected with computing machine through video transmission line.The particular location that camera sets up, quantity and height are confirmed according to the concrete condition in tested waters.In order to the water surface situation of monitoring tested waters (changing the most significantly because of the breakout of water bloom in waters is the variation of water surface visual signature) and absorb the video image of the water surface, reach the purpose of real-time monitoring Surface Picture.

2. back projection is handled

After the completion of (2)-1. step; Each frame video image to the input of (2)-1. step carries out back projection processing (Back Projection); Promptly to each pixel in the video image processing region, the matching degree of the color prior model of setting up through this pixel of computer inquery and (1) step (being H component tone probability model), (other the regional probability outside this zone are 0 just to obtain the probability that this pixel is an object pixel; The value of each pixel has just become a kind of discretize tolerance that color of object information appears at the possibility here in the image; The possibility that here occurs is big more, and the value of pixel is just big more, otherwise then more little).Through above-mentioned processing, just obtain the color of object back projection figure of every two field picture.

3. detect and cut apart the wawter bloom color lump and filter the interference color lump

(2)-2. step was detected the wawter bloom color lump and disturbs color lump to handle with filtering, that is: after accomplishing

The back projection figure that earlier (2)-2. step was obtained carries out burn into and expands, and detects the inside and outside contour of target color lump and finds cavity in the target color lump to fill and wait pre-service, so just can remove a part of interference color lump, avoids noise to a certain extent.

Detect again and cut apart the wawter bloom color block areas, promptly press following formula constantly to the back projection figure computation of mean values translation vector M after the above-mentioned pre-service of process through computing machine _{H, G (x)}:

$M_{h,G\left(x\right)}=\frac{{\mathrm{<figure-callout\; id="2"\; label="h"\; filenames="HSA00000171985400011.png"\; state="\{\{state\}\}">h</figure-callout>}}^2\&dtri;{\hat{f}}_K\left(x\right)}{2/C{\hat{f}}_G\left(x\right)}---\left(2\right)$

Wherein: h is a bandwidth; C is a normaliztion constant,

be that the last bandwidth of nuclear K is the multivariate density Estimation of h:

${\hat{f}}_K\left(x\right)=\frac{1}{n{h}^d}\underset{j=1}{\overset{n}{\mathrm{\&Sigma;}}}k\left({\left|\right|\frac{x-x_i}{h}\left|\right|}^2\right)---\left(3\right)$

Set x ' _j} _{J=1,2 ... N}Be d dimension Euclidean space R ^dN point, k (x) representes the kernel function of this pixel:

Wherein: c _dBe d dimension unit spheroid volume.In order to the position of iteration renewal target color lump, up to the Optimum Matching point that converges on, the zone of this match point is exactly the target color lump (being the wawter bloom color block areas) in the current frame image.

Judge: when not detecting the target color lump, think that then wawter bloom does not appear in the water surface, returned for (2)-1. step and obtain next frame and proceed to detect; When detecting the target color lump, then with the barycenter of present frame Optimum Matching point and area as the initial value of next frame search window, proceed iterative computation and find the Optimum Matching point.So till loop iteration to the EOP (end of program), just detect the target color lump that is partitioned into every two field picture.

Filter then and disturb color lump; The area of the present frame that the last step of i.e. calculating obtains and the target color lump of next frame; Interframe Changing Pattern according to the wawter bloom region area; Utilize the target color lump area of present frame to predict the target color lump area in the next frame image, promptly calculate the difference (being redundance) between next frame color lump actual detected area and the next frame prediction area, judge according to following formula:

|S _n+1-S′ _n+1|≤ξ (5)

Wherein: S _N+1Be the actual measurement area of next frame color lump, S ' _N+1Be the prediction area of next frame, the error range of ξ for allowing.

When | S _N+1-S ' _N+1| during≤ξ, then think the next frame actual detected to the color lump area meet the interframe Changing Pattern of wawter bloom region area and with it as testing result; When | S _N+1-S ' _N+1|＞ξ, when promptly difference exceeds in the error range of permission, then think the next frame actual detected to the color lump area do not meet the Changing Pattern of interframe wawter bloom region area, redundance is defined as disturbs color lump and it is filtered.

Through after the above-mentioned processing, just detect to be partitioned into and filter the wawter bloom color lump that disturbs behind the color lump, obtain the elemental area of wawter bloom color lump in image.Then testing result is demarcated and exported to the wawter bloom color block areas in every two field picture.

(3) bloom prealarming

1. calculate wawter bloom zone real area:

(2) step calculated (2) earlier and goes on foot the number percent that the elemental area of wawter bloom color lump in image that obtains accounts for the visual field pixel total area after accomplishing, and the actual total area that multiply by the visual field just can obtain wawter bloom zone real area.Pass through camera calibration again; Calculate the actual total area in the visual field; The object of reference that is about to known area places the position, waters under the camera, calculates the ratio that its elemental area in image accounts for the visual field pixel total area, just obtains the actual total area in the visual field divided by this ratio with the real area of this known object of reference; And the wawter bloom that obtains zone real area outputed on the computer software interface, so that the user gets information about the real area in wawter bloom zone.

2. set up Early-warning Model:

After the completion of (3)-1. step; Go on foot the wawter bloom zone real area that obtains according to (3)-1.; Be at interval by 8 hours earlier, calculate the average of the wawter bloom zone real area in each time interval, set up the one-variable linear regression forecast model through the Change in Mean rule of this real area of Computer Analysis again; Predict the average of the real area in the next time interval; And the average of calculating this real area accounts for the number percent of the waters total area, and the wawter bloom disaster degree with this number percent representative is a foundation then, and the advanced warning grade of breakout of water bloom is divided into blueness, yellow, orange and red four grades; Corresponding early warning interval be respectively generally, heavier, serious and four grades of especially severe; And judge:, then turn back to and returned for (2)-1. step and obtain next frame and proceed detection when the number percent of the wawter bloom disaster degree of representative during less than early warning value (size of early warning value confirms that according to the concrete condition in tested waters and application demand the present invention gets 5%); When the number percent of wawter bloom disaster degree of representative during, then carry out early warning according to the advanced warning grade of dividing greater than early warning value.The concrete division of breakout of water bloom advanced warning grade is following:

General blue early warning (being fragmentary property wawter bloom): wawter bloom is sporadicly gathered, and main waters district algae bio density is less than 3,000 ten thousand/L, and the wawter bloom area is more than or equal to 5% of the water body total area.

Heavier yellow early warning (being the locality wawter bloom): algae gathers in local waters, and main waters district algae bio density is between 3000～5,000 ten thousand/L, and the wawter bloom area is more than or equal to 10% of the water body total area.

Serious orange early warning (being regional wawter bloom): when generation area property wawter bloom, main waters district algae bio density is between 5000～8,000 ten thousand/L, and the wawter bloom area is more than or equal to 40% of the water body total area.

Especially severe red early warning (being comprehensive wawter bloom): wawter bloom is comprehensive to be broken out, and the algae bio density in main waters district is greater than 8,000 ten thousand/L, and the wawter bloom area is more than or equal to 60% of the water body total area.

After this method for monitoring and early warning water bloom in small water area based on Flame Image Process tested, can obtain to draw a conclusion:

1. in observation process; Tested waters (Chongqing City splash Kai Xian small stream River Reservoir) water surface generation wawter bloom; The inventive method can detect the wawter bloom color block areas in real time and calculate this regional real area, can accurately carry out early warning according to advanced warning grade, at breakout of water bloom in earlier stage; Chongqing City Environmental Protection Agency can in time grasp the wawter bloom the condition of a disaster in this waters and notify relevant departments to take corresponding measure to administer (when fragmentary property wawter bloom, salvaging), strong cooperation the work of preventing and reducing natural disasters.

2. the inventive method only use a computer with a camera just can realize the wawter bloom monitoring and warning fast and effectively; Practiced thrift the monitoring cost to a great extent, and the user is easy to use, maintenance cost is low; Monitoring effect can satisfy the actual demand of Chongqing City Environmental Protection Agency, thereby easy to utilize.

3. the present invention adopts improved continuous adaptive property average drifting algorithm to detect to cut apart the wawter bloom color lump and filters and disturb color lump, can effectively suppress the noise that water surface background complicacy and ambient light variation etc. bring.This algorithm computation amount is little, speed is fast, is the basis that guarantees wawter bloom monitoring and warning real-time.

The monitoring and early warning water bloom in small water area based on Flame Image Process that above-mentioned conclusion explanation utilizes the inventive method to realize can be realized the real-time monitoring and warning of wawter bloom; Monitoring result meets the actual distribution scope in wawter bloom zone; Monitoring effect can satisfy the actual demand of use department, and the monitoring cost is low, and is strong to adaptive capacity to environment; Easy to utilize, so the present invention can be applied in the actual project.

Claims (1)

1. the method for monitoring and early warning water bloom in small water area based on Flame Image Process is characterized in that utilizing camera and computing machine, calculates through program, and its concrete steps are following:

(1) sets up the color prior model

At first, set up the color prior model, that is: according to the historical wawter bloom in tested waters regional sample figure or video data

If x ' _j} _{J=1,2 ... N}Be n pixel of sample figure or video data, with each pixel from the three primary colors space conversion to look, show the score from the space, be i with the discretize value of tone component, i.e. i=1,2..., 360, zoom to [0,255] to the scope of tone component [0,360]; Be calculated as follows the tone probability of tone component through computing machine:

p＝{p _i} _{i＝1，2…m}； $p_i=\frac{\underset{j=1}{\overset{n}{\mathrm{\&Sigma;}}}{\left|\right|x_j^{\&prime;}\left|\right|}^2\mathrm{\&delta;}[b\left(x_j^{\&prime;}\right)-i]}{\underset{j=1}{\overset{n}{\mathrm{\&Sigma;}}}\left({\left|\right|x_j^{\&prime;}\left|\right|}^2\right)};$ $\underset{i=1}{\overset{m}{\mathrm{\&Sigma;}}}{p}_i=1---\left(1\right)$

Wherein: p is a color model, p _iBe that tone value is the probability of i, δ is the δ function, function b (x ' _j) be space R ²→ { 1,2 ... The index of m} promptly is positioned at position x ' _jPixel to the index in histogram quantization characteristic space;

(2) the wawter bloom zone is detected

1. small-sized waters is monitored

(1) step was set up camera on the bank in small-sized waters after accomplishing, and was connected with computing machine through video transmission line; The particular location that camera sets up, quantity and height are confirmed according to the concrete condition in tested waters;

2. back projection is handled

After the completion of (2)-1. step; Each frame video image to the input of (2)-1. step carries out the back projection processing; Promptly to each pixel in the video image processing region; The matching degree of the color prior model of setting up through this pixel of computer inquery and (1) step just obtains the color of object back projection figure of every two field picture;

3. detect and cut apart the wawter bloom color lump and filter the interference color lump

(2)-2. step was detected the wawter bloom color lump and disturbs color lump to handle with filtering, that is: after accomplishing

The back projection figure that earlier (2)-2. step was obtained carries out burn into and expands, and detects the inside and outside contour of target color lump and finds the cavity in the target color lump to fill pre-service;

Detect again and cut apart the wawter bloom color block areas, promptly press following formula constantly to the back projection figure computation of mean values translation vector M after the above-mentioned pre-service of process through computing machine _{H, G (x)}:

$M_{h,G\left(x\right)}=\frac{h^2\&dtri;{\hat{f}}_K\left(x\right)}{2/C{\hat{f}}_G\left(x\right)}---\left(2\right)$

Wherein: h is a bandwidth; C is a normaliztion constant,

be that the last bandwidth of nuclear K is the multivariate density Estimation of h:

${\hat{f}}_K\left(x\right)=\frac{1}{n{h}^d}\underset{j=1}{\overset{n}{\mathrm{\&Sigma;}}}k\left({\left|\right|\frac{x-x_i}{h}\left|\right|}^2\right)---\left(3\right)$

Set x ' _j} _{J=1,2 ... N}Be d dimension Euclidean space R ^dN point, k (x) representes the kernel function of this pixel:

Wherein: c _dBe d dimension unit spheroid volume; In order to the position of iteration renewal target color lump, up to the Optimum Matching point that converges on, the zone of this match point is exactly the target color lump in the current frame image;

Judge: when not detecting the target color lump, then returned for (2)-1. step and obtain next frame and proceed to detect; When detecting the target color lump; Then with the barycenter of present frame Optimum Matching point and area initial value as the next frame search window; Proceed iterative computation and find the Optimum Matching point, so till loop iteration to the EOP (end of program), just detect the target color lump that is partitioned into every two field picture;

Filter then and disturb color lump; The area of the present frame that the last step of i.e. calculating obtains and the target color lump of next frame; Interframe Changing Pattern according to the wawter bloom region area; Utilize the target color lump area of present frame to predict the target color lump area in the next frame image, promptly calculate the difference between next frame color lump actual detected area and the next frame prediction area, judge according to following formula:

|S _n+1-S′ _n+1|≤ξ (5)

Wherein: S _N+1Be the actual measurement area of next frame color lump, S ' _N+1Be the prediction area of next frame, the error range of ξ for allowing;

When | S _N+1-S ' _N+1| during≤ξ, then think the next frame actual detected to the color lump area meet the interframe Changing Pattern of wawter bloom region area and with it as testing result; When | S _N+1-S ' _N+1|＞ξ, when promptly difference exceeds in the error range of permission, then think the next frame actual detected to the color lump area do not meet the Changing Pattern of interframe wawter bloom region area, redundance is defined as disturbs color lump and it is filtered; Detection is partitioned into and filters the wawter bloom color lump that disturbs behind the color lump fixed and output testing result of rower of going forward side by side;

(3) bloom prealarming

1. calculate wawter bloom zone real area:

(2) step calculated (2) earlier and goes on foot the number percent that the elemental area of wawter bloom color lump in image that obtains accounts for the visual field pixel total area after accomplishing, and the actual total area that multiply by the visual field just can obtain wawter bloom zone real area; Pass through camera calibration again; Calculate the actual total area in the visual field; The object of reference that is about to known area places the position, waters under the camera; Calculate the ratio that its elemental area in image accounts for the visual field pixel total area, just obtain the actual total area in the visual field divided by this ratio, and the wawter bloom zone real area that obtains is outputed on the computer software interface with the real area of this known object of reference;

2. set up Early-warning Model:

After the completion of (3)-1. step; Go on foot the wawter bloom zone real area that obtains according to (3)-1.; Be at interval by 6～10 hours earlier; Calculate the average of the wawter bloom zone real area in each time interval, set up the one-variable linear regression forecast model through the Change in Mean rule of this real area of Computer Analysis again, predict the average of the real area in the next time interval; And the average of calculating this real area accounts for the number percent of the waters total area; Wawter bloom disaster degree with the representative of this number percent is a foundation then, and the advanced warning grade of breakout of water bloom is divided into blueness, yellow, orange and red four grades, corresponding early warning interval be respectively generally, heavier, serious and four grades of especially severe; And judge:, then turn back to and returned for (2)-1. step and obtain next frame and proceed to detect when the number percent of wawter bloom disaster degree of representative during less than early warning value; When the number percent of wawter bloom disaster degree of representative during, then carry out early warning according to the advanced warning grade of dividing greater than early warning value;

The concrete division of breakout of water bloom advanced warning grade is following:

General blue early warning: wawter bloom is sporadicly gathered, and main waters district algae bio density is less than 3,000 ten thousand/L, and the wawter bloom area is more than or equal to 5% of the water body total area;

Heavier yellow early warning: algae gathers in local waters, and main waters district algae bio density is between 3000～5,000 ten thousand/L, and the wawter bloom area is more than or equal to 10% of the water body total area;

Serious orange early warning: when generation area property wawter bloom, main waters district algae bio density is between 5000～8,000 ten thousand/L, and the wawter bloom area is more than or equal to 40% of the water body total area;

The especially severe red early warning: wawter bloom is comprehensive to be broken out, and the algae bio density in main waters district is greater than 8,000 ten thousand/L, and the wawter bloom area is more than or equal to 60% of the water body total area.

Images