CN106600541A - multi-mode transmission video image sharpening processing system based on self-adaptive atmospheric light curtain graph - Google Patents

Abstract

The invention discloses a multi-mode transmission video image sharpening processing system based on a self-adaptive atmospheric light curtain graph. The multi-mode transmission video image sharpening processing system comprises front-end video image acquisition equipment, a multi-mode-based remote transmission module and an image processing service terminal, wherein a carrying platform comprises an unmanned aerial vehicle, an unmanned ship, a fixed end and hand-held mobile camera shooting equipment; the unmanned aerial vehicle and the unmanned ship are used for taking pictures and video of a required scene to obtain a desired image source through controlling relevant equipment remotely in the filed by operators; the fixed end just needs to control the camera shooting equipment to acquire images remotely; the hand-held mobile camera shooting equipment needs to be operated by the staff to conduct specific image acquisition according to actual needs; and a server side is used for carrying out sharpening processing on the received images, and an image defogging algorithm of the self-adaptive atmospheric light curtain graph is run at the server side, thereby realizing the effects of defogging and sharpening quickly and effectively. The multi-mode transmission video image sharpening processing system can realize multi-scene image acquisition, and can realize simple, fast and real-time defogging to obtain clear images.

Description

Multimode transmissions video image sharpening processing system based on self adaptation air light curtain figure

Technical field

The present invention relates to terminal taking, is wirelessly transferred and intelligent image process field, more particularly to it is big based on self adaptation The multimode transmissions video image sharpening processing system of gas light curtain figure.

Background technology

Due to being affected by vapor and dust in air etc., its definition is greatly reduced outdoor imaging system, Image integrally whitens, textural characteristics, and detailed information is lost.In addition, with industrialized progress faster, atmospheric pollution is increasingly tight Weight, haze weather is frequently occurred so that atmospheric visibility is strongly reduced, and the problem being accompanied by causes outdoor imaging more difficult, room Needed for the image definition of outer acquisition does not much reach the requirement required for intelligent use system and the daily outdoor of people is taken pictures.Drop Map picture loses the value of its application significantly, therefore image mist elimination sharpening technology has critically important Research Significance.

Current image mist elimination sharpening technology is broadly divided into two classes：Traditional image procossing skill is relied on using image enhaucament Art, its technology maturation, but the method can only be directed to picture contrast or rectangular histogram and carry out mist elimination, not consider the texture letter of image Often there is the problems such as effective information is lost, textural characteristics are lacked in breath, its result.Another kind of is based on atmospheric scattering thing The method of reason model, the method are according to the degeneration principle for having mist image, by the physical model for setting up atmospheric scattering effect, profit Solution atmospherical scattering model is carried out with addition prior information, so as to obtain clear fogless image, the method is based on image-forming principle, Its treatment effect is more excellent, effective information with and textural characteristics recover more intact.

In view of traditional mist elimination technology exists computationally intensive, operation efficiency is low, can not be automatic to the region of different mistiness degree The different degrees of limitation such as mist elimination corrected parameter are set, and the present invention discloses a kind of self adaptation mist elimination side based on air light curtain Method, based on atmospherical scattering model and is improved, and has mist Image estimation prior information by original, and substituting into atmospherical scattering model is carried out Solve clear fogless image.The method realizes that to the fog-zone domain of image variable concentrations self adaptation arranges mist elimination parameter so that go The result entirety visual effect of mist sharpening is more preferably.Described self adaptation mist elimination algorithm complex is low, and operational efficiency is high, adapts to Can only image processing system in real time.

The content of the invention

The described multimode transmissions video image sharpening processing system based on self adaptation air light curtain figure, its feature exist In, including：

The acquisition of multiterminal image, using UAV flight's video camera, unmanned boat carries video camera, and hard-wired monitoring sets The equipment such as standby and hand-held mobile video camera carry out obtaining for image to different scenes such as farmland, building, traffic and oceans Take；

Multiple terminals data transfer, in multiple terminals using 4G radio network gateways by the data back for getting to server, 4G nets Cover a wide range, transmission speed is fast, using with countryside or the scene for being difficult to wire transmission.

Server end, it is main to realize the view data of multiple terminals collection being carried out to receive storage and being carried out based on air light curtain Adapting to image sharpening computing.

Described multiterminal transmission Sharp processing of image system, it is characterised in that：Described server end carries out data and connects Receive and computing.

Described multiterminal transmission Sharp processing of image system, it is characterised in that：At described self adaptation mist elimination sharpening Adjustment method includes received image is carried out asking for the minima image W of its tri- Color Channel of R, G, B₀, wherein W₀Definition For：

Described multiterminal transmission Sharp processing of image system, it is characterised in that：At described self adaptation mist elimination sharpening Adjustment method includes that carrying out Laplce's filtering operation with described minima image extracts texture feature information therein, and from Filter in minima image, obtain described air light curtain image V (x), its mathematic(al) representation is：

Wherein H is Laplace operator.

Described multiterminal transmission Sharp processing of image system, it is characterised in that：Described air light curtain figure is directly carried out Solve picture rich in detail and can not obtain satisfied image mist elimination sharpening result, it is therefore desirable to add adaptive correction ginseng to which Number, wherein adaptive correction parameter is tried to achieve using the minima image local mean square error, and its mathematic(al) representation is：

Described multiterminal transmission Sharp processing of image system, it is characterised in that：At described self adaptation mist elimination sharpening Adjustment method includes the solution that air light value figure is carried out to original image, and its calculating process is that the luminance picture to original image is carried out Medium filtering filters textural characteristics present in luminance picture；As medium filtering causes filtered image edge distortion, therefore this Invention carries out edge reconstruction to filtered image using Steerable filter.Through described filtering twice, it is of the present invention from Adapt to mist algorithm for image clearness and try to achieve more accurate air light value figure A (x).

Described multiterminal transmission Sharp processing of image system, it is characterised in that：At described self adaptation mist elimination sharpening Adjustment method is to solve prior information including utilization described according to the characteristic of the original image based on atmospherical scattering model Air light curtain figure, described adaptive correction parameter and described air light value figure；Then described prior information is substituted into and is dissipated Penetrating model carries out solving fogless image, and its solution mathematic(al) representation is：

It is of the invention compared with traditional mist elimination sharpening technology, with advantages below and beneficial effect：

1st, the present invention realizes that the integrated of many scene imaging systems enables mist elimination sharpening processing system of the present invention Enough it is adapted to and applies in various scenes.

2nd, multiterminal transmission Sharp processing of image system of the present invention carries out high in the clouds transmission using 4G radio network gateways, its Possess wide coverage, transmission speed is fast, it is not necessary to draw in the net line so that its cost is substantially reduced, and solve some areas without Limitation of the method using wired network

3rd, multiterminal of the present invention transmit Sharp processing of image system, using data back, the side of server process Formula carries out image procossing.Without the need for exclusive equipment, the cost of capture apparatus is reduced, and server operation speed is fast and has There is stability.

4th, multiterminal of the present invention transmit Sharp processing of image system, disclose one kind and are based on air light curtain image certainly Adapt to mist elimination algorithm.Its algorithm complex is low, and operation efficiency is high, can realize adaptive function to image mist elimination so that go After mist sharpening, image entirety visual effect and information retain more preferably.

Description of the drawings

Fig. 1 is multiterminal of the present invention based on the multimode transmissions video image sharpening processing system of self adaptation air light curtain figure Image transmitting schematic diagram；

Fig. 2 is the present invention based on the clear of the multimode transmissions video image sharpening processing system of self adaptation air light curtain figure Change algorithm workflow diagram；

Fig. 3 is that multimode transmissions video image sharpening processing system of the present invention based on self adaptation air light curtain figure processes institute State the degraded image sharpening effect contrast figure of acquisition.

Specific embodiment

In order that the objects, technical solutions and advantages of the present invention become more apparent, below in conjunction with accompanying drawing and example, to this Invention is further elaborated.

Such as Fig. 1, a kind of many scene image acquisition methods in multiple terminals of the technical program offer, the degraded image acquisition methods Corresponding image is obtained in real time for misty rain weather and the image to the acquisition carries out adapting to image of the present invention and goes Mist sharpening computing, so as to obtain clear fogless image, described mist elimination image acquisition methods include step：

The many different scenes of real-time control photographic head carry out image acquisition according to specific needs, and described photographic head includes nobody Machine carries video camera, and unmanned boat carries video camera, fixedly mounts video camera and hand-held mobile video camera etc., it is adaptable to various The image acquisition of scene.

Described multiple terminals is passed to server transport via high in the clouds in real time by 4G radio network gateways after collecting image Transmission of data.

Described server possesses the data collected via the multiple terminals that high in the clouds transmits by reception, and to the end The original image that end obtains carries out the process of image mist elimination sharpening, to obtain clearly image.

Described mist elimination sharpening Processing Algorithm flow process is as shown in Fig. 2 specific as follows：

The first step, original image I (x) received according to the server ask for the minima of tri- Color Channel of R, G, B Image W₀, i.e., compare the worth size of its three Color Channels to each pixel in original image, by minimum therein Value is stored in minima image W₀, wherein W₀The mathematic(al) representation of definition is：

Second step, according to described minima image W₀Carry out asking for for air light curtain image.According to atmospherical scattering model And the definition of air light curtain image, air light curtain figure do not have correlation, wherein minima image W with image unity and coherence in writing feature₀Comprising Abundant textural characteristics, therefore is needed to be filtered which texture information is filtered.The technology that the present invention is used is to use drawing general Lars filter operator is filtered process to minima image, and which can effectively filter out the textural characteristics of minima image and have Good marginal information, obtains ideal air light curtain figure V (x).

3rd step, described air light curtain image can not realize self adaptation mist elimination function well, and the present invention is to air Light curtain figure adds corrected parameter so as to adaptive function.Wherein the present invention asks for local mean square error to minima image Differ from, and define its auto-adaptive parameter and be：

Wherein x²The mean square error being defined as in a certain regional area, σ²It is defined as the ginseng for Gaussian function Number, this algorithm value are 0.5.

4th step, the present invention remain a need for asking for air light value image to realize subsequent step to institute according to the original image Stating original image carries out mist elimination sharpening process.Luminance picture I of the present invention to described original image_vCarry out medium filtering reality Now the smoothing processing to luminance picture, subsequently carries out edge reconstruction to the image after the smoothing processing using Steerable filter, with The air light value figure required for the present invention is obtained, wherein luminance picture is defined as：

C represents R in coloured image, some passage of G, B.

5th step, the present invention are based on according to atmospherical scattering model and based on one kind disclosed in air light curtain mist elimination technology The adapting to image mist elimination technology of air light curtain.The present invention utilizes described air light curtain image V (x), auto-adaptive parameter P (x) And air light value image A (x) carries out asking for clear fogless image, its mathematic(al) representation is：

Multimode transmissions video image sharpening processing system based on self adaptation air light curtain figure of the present invention is to described The degraded image of acquisition carries out the effect of mist elimination sharpening as shown in figure 3, wherein Fig. 3 (a) is the degraded image of the acquisition, figure The multimode transmissions video image sharpening processing system of 3 (b) described in this patent based on self adaptation air light curtain figure is to the figure that degrades The mist elimination sharpening result of picture.At the multimode transmissions video image sharpening based on self adaptation air light curtain figure described in this patent Reason system has good sharpening treatment effect to image containing mist, and sharpening processing system of the present invention has place Reason real-time, it is adaptable among adapting to production application.

Above the specific embodiment of the present invention is described.It is to be appreciated that the invention is not limited in above-mentioned Particular implementation, those skilled in the art can make various modifications or modification within the scope of the claims, this not shadow Ring the flesh and blood of the present invention.

Claims (6)

1. a kind of multimode transmissions video image sharpening processing system based on self adaptation air light curtain figure, is characterised by, this is System includes：

Multiterminal image is obtained, and the acquisition of multiterminal is carried out to different scenes image using equipment；

Multiple terminals data transfer, carries out being wirelessly transferred for image using 4G radio nodes gateway, the image that different terminals are obtained Data use wireless network transmissions, are received in server end；

Server end, receives to the view data that the multiterminal image is obtained, and runs based on the adaptive of air light curtain figure Answer mist elimination algoritic module to obtain clear fogless image, guarantee is provided for successive image process work.

2. the multimode transmissions video image sharpening processing system based on self adaptation air light curtain figure according to claim 1 System, it is to carry video camera, hard-wired prison by using UAV flight's video camera, unmanned boat that described multiterminal image is obtained Control equipment or hand-held mobile video camera carry out the acquisition of image to the different scenes in farmland, building or lake class.

3. the multimode transmissions video image sharpening processing system based on self adaptation air light curtain figure according to claim 2 System, it is characterised in that UAV flight's video camera, unmanned boat carry shooting etc. equipment be connected with 4G radio network gateways, with The function of data is sent, the image for getting can be wirelessly transferred.

4. the multimode transmissions video image sharpening processing system based on self adaptation air light curtain figure according to claim 1 System, it is characterised in that：

Characterized in that, the server end possesses the view data for receiving the multiple terminals data transfer passback, and to being obtained The view data for taking carries out the self adaptation mist elimination computing based on air light curtain, realizes image sharpening function.

5. the multimode transmissions video image sharpening processing system based on self adaptation air light curtain figure according to claim 4 System, it is characterised in that：

Described self adaptation mist elimination algoritic module includes：

Original image I (x) received according to the server asks for the minima image W of tri- Color Channel of R, G, B₀, i.e., to original Each pixel in beginning image compares the worth size of its three Color Channels, and minima therein is stored in minima figure As W₀, wherein W₀The mathematic(al) representation of definition is：

$W_0=\underset{C\∈\left\{R,G,B\right\}}{\min }{I}^C\left(x\right);$

According to described minima image W₀Asking for for air light curtain image is carried out, to minima image W₀It is filtered texture Information Filtration, is filtered process to minima image using Laplce's filter operator, and which can effectively filter out minima image Textural characteristics and have good marginal information, obtain ideal air light curtain figure V (x)；

Using described minima image W₀Carry out asking for adaptive correction parameter P (x), wherein the method asked for is using minimum It is worth the local mean square error of image, adaptive correction parameter P (x) is：

$P\left(x\right)=e^{-\frac{x^2}{2{\mathrm{\σ}}^2}};$

Wherein, x2 is defined as the mean square error in a certain regional area, σ²It is defined as the parameter for Gaussian function；

Air light value image is asked for according to the original image, the luminance picture I to described original image_vX () carries out intermediate value filter Ripple realizes the smoothing processing to luminance picture, subsequently carries out edge weight to the image after the smoothing processing using Steerable filter Build, to obtain air light value figure A (x), luminance picture I_vX () is defined as：

$I_v\left(x\right)=\underset{C\∈\{R,G,B\}}{max}{I}^C\left(x\right);$

Wherein, C represents R in coloured image, some passage of G, B；

By described original image I (x), air light curtain figure V (x), adaptive correction parameter P (x) and air light value figure A (x) Substitute into atmospherical scattering model and ask for clear fog free images, its mathematic(al) representation is as follows：

$J=\frac{I\left(x\right)-V\left(x\right)*P\left(x\right)}{1-V\left(x\right)*P\left(x\right)/A\left(x\right)}.$

6. the multimode transmissions video image sharpening processing system based on self adaptation air light curtain figure according to claim 5 System,

It is characterized in that：The expression formula of described air light curtain figure is：Wherein H is La Pula This operator.