CN104408709A - Rapid correction method of image with non-uniform gray scale of linear-array CCD camera - Google Patents

Abstract

The invention relates to a rapid correction method of an image with a non-uniform gray scale of a linear-array CCD camera. The linear-array CCD camera collects an image of a to-be-measured object, the image is corrected line by line, and then a corrected image is formed by splicing and combination; during each line scanning period, gray levels of all pixels of a line image of the to-be-measured object image are arranged at a corresponding linear-array CCD pixel one-dimensional direction to form gray level distribution function; and correction is carried out based on a population-mean-based moving average idea. According to the invention, the correction method has the high real-time performance. Defects of high cost, low precision, and short service life and the like of hardware correction as well as defects of complex algorithm, poor real-time performance, requirement of complex mathematic model establishment, and requirement of parameter calibration of the traditional software correction method can be overcome. Therefore, with comprehensive consideration of the total gray mean value and the local gray mean value, gray compensation and correction of the dark parts of two sides of the CCD imaging are realized, thereby realizing real-time gray correction of the linear-array CCD.

Description

The irregular quick antidote of a kind of linear array CCD camera gradation of image

Technical field

The invention belongs to technical field of image processing, relate to the irregular quick antidote of a kind of linear array CCD camera gradation of image.

Background technology

CCD (Charge Couple Device) camera is divided into face battle array and linear array two kinds, and wherein linear array CCD camera is very suitable for the image acquisition of high-speed moving object, is now widely used in the fields such as Machine Vision Detection.But under the demand of high precision and Large visual angle image acquisition, the imaging pixel number of linear array CCD camera is also increase thereupon, the loss that directly results in luminous energy arrival imaging plane significantly increases, make the gray scale of the output image of CCD camera there is the dark irregular phenomenon of illumination in middle bright both sides, seriously have impact on the operations such as follow-up image analysis processing.Therefore irregular to the gray scale exported image carries out corresponding corrective operations and seems very necessary.

In actual applications, as shown in Figure 1, linear array CCD camera imaging pixel arranges along straight line, and due to optical energy loss, the optical axis place energy of the imaging plane of optical imaging system is maximum, and both sides light energy losses is larger.Now, even if adopt parallel uniform source of light to be radiated on measured object, the light intensity arrived on imaging plane also there will be the phenomenon pockety because of optical energy loss, causes collected image intensity value to present the darker irregular result in middle bright both sides, as shown in Figure 2.

At present, the antidote irregular for linear array CCD camera output image gray scale is divided into two kinds: (1) hardware is corrected: carry out sectional-regulated to illumination system, the LED light source generally selecting sectional to regulate carries out illumination compensation to two ends, though this method does not increase the excessive data processing time, there is Automatic adjusument difference and the precision deficiency such as low; (2) software is corrected: mainly contain histogram equalization, based on frequency domain filtering and with reference to antidotes such as standard specimens, these class methods need to set up complicated mathematical model usually, its correlation parameter also needs to demarcate in advance, is difficult to meet the demand such as high real-time and self-adaptation in practical application.

Summary of the invention

The present invention is that the deficiency overcoming the existence of above-mentioned prior art proposes the irregular quick antidote of a kind of linear array CCD camera gradation of image, the method is simple to operate, adaptivity good, without the need to setting up complicated mathematical model and parameter calibration operation, can the intensity profile of efficient, flexible ground orthotic line array CCD camera image irregular.The present invention exports the imaging law of a line view data combining linear array CCD camera each scan period, by the real-time process to progressive image data, propose a kind of irregular antidote of real-time gray scale in conjunction with population mean and the average thought of micro-slip.

The irregular quick antidote of a kind of linear array CCD camera gradation of image of the present invention, it is the good antidote of a kind of real-time, overcoming algorithm complexity, poor real that the high cost that hardware corrects, low precision and the life-span not enough and traditional software antidote such as low exists, complicate mathematical model need set up and carry out on the shortcoming bases such as parameter calibration, by comprehensively taking into account overall gray level average and local gray average, realize correcting the grey level compensation of CCD imaging both sides compared with dark-part, thus achieve the real-time gray-level registration of line array CCD.

The quick antidote that linear array CCD camera imaging gray scale is irregular, linear array CCD camera gathers measured object image, corrects this image line by line, then pieces together the image after rectification; Each line period, the gray-scale value of each pixel of row image of measured object image presses the one-dimensional square of the pixel of corresponding line array CCD to arrangement, forms a grey value profile function, is designated as F (x); Because optical energy loss imaging obviously can present the dark irregular phenomenon of illumination in middle bright both sides, need correct F (x), obtaining the grey value profile function after correcting is F ' (x), and its concrete steps of correcting are as follows:

(1) solve the overall gray level mean value of scan line image corresponding to F (x), and be designated as G

$G=\frac{\underset{x=1}{\overset{N}{\mathrm{\Σ}}}F\left(x\right)}{N};$

Wherein, N is the number of pixels of current line image;

(2) F (x) is solved in interval sliding average, and be designated as M (x)

$M\left(x\right)=\frac{{\mathrm{\Σ}}_{x-m}^{x+n}F\left(i\right)}{m+n+1};$

Wherein, m is forward slip step-length, and n slides backward step-length, total step-length step=m+n+1,0<step<N/4; I is temporary variable,

(3) the grey value profile function F after obtaining rectification is solved ' (x)

F’(x)＝G+(F(x)-M(x))。

As above the irregular quick antidote of a kind of linear array CCD camera gradation of image described in 1, m+1=n.

The irregular quick antidote of a kind of linear array CCD camera gradation of image as above, pixel number >=1024 of linear array CCD camera.

Beneficial effect

1, method calculating is simple, real-time is good;

2, compared to the hardware type antidote that traditional employing additional light source compensates, described method has that self-adaptation is good, precision is high and without advantages such as restrictions in serviceable life;

3, method takes into account the inner link of population mean and micro-slip mean value, and its Output rusults has very strong adaptivity.

Accompanying drawing explanation

Fig. 1 is line array CCD optical energy loss schematic diagram

Fig. 2 is the irregular schematic diagram of line array CCD imaging intensity profile

Fig. 3 is the schematic diagram of running mean thought for correcting that the present invention is based on population mean

Fig. 4 is web original image

Web image after correcting when Fig. 5 is step-length step=192

Fig. 6 is the irregular comparison diagram of intensity profile before and after correcting

Embodiment

Below in conjunction with embodiment, set forth the present invention further.Should be understood that these embodiments are only not used in for illustration of the present invention to limit the scope of the invention.In addition should be understood that those skilled in the art can make various changes or modifications the present invention, and these equivalent form of values fall within the application's appended claims limited range equally after the content of having read the present invention's instruction.

The quick antidote that a kind of linear array CCD camera imaging gray scale is irregular, linear array CCD camera gathers measured object image, correct this image line by line, then piece together the image after rectification, Fig. 3 is the schematic diagram of running mean thought for correcting that the present invention is based on population mean; Each line period, the gray-scale value of each pixel of row image of measured object image presses the one-dimensional square of the pixel of corresponding line array CCD to arrangement, forms a grey value profile function, is designated as F (x); Because optical energy loss imaging obviously can present the dark irregular phenomenon of illumination in middle bright both sides, affect image quality, need correct F (x), obtaining the grey value profile function after correcting is F ' (x), and its concrete steps of correcting are as follows:

(1) solve the overall gray level mean value of scan line image corresponding to F (x), and be designated as G

$G=\frac{\underset{x=1}{\overset{N}{\mathrm{\&Sigma;}}}F\left(x\right)}{N};$

Wherein, N is the number of pixels of current line image;

(2) F (x) is solved in interval sliding average, and be designated as M (x)

$M\left(x\right)=\frac{{\mathrm{\&Sigma;}}_{x-m}^{x+n}F\left(i\right)}{m+n+1};$

Wherein, m is forward slip step-length, and n slides backward step-length, total step-length step=m+n+1,0<step<N/4; I is temporary variable,

(3) the grey value profile function F after obtaining rectification is solved ' (x)

F’(x)＝G+(F(x)-M(x))。

The irregular quick antidote of a kind of linear array CCD camera gradation of image as above, m+1=n.

The irregular quick antidote of a kind of linear array CCD camera gradation of image as above, pixel number >=1024 of linear array CCD camera.

Embodiment 1

Fig. 4 does not utilize the method to carry out the web original image corrected, and this image is in middle bright as can be seen from this figure, and the irregular distribution characteristics of gray scale that both sides are dark, now collected picture quality is poor.Utilize the method to be corrected below, be that the row image progressive that CCD in each line period collects is corrected in practical application, then piece together a width complete image.Here for be contrasted with former figure, first split line by line by the web original image after collection, then correct line by line, finally piece together a width complete image again, concrete steps are as follows:

The gray-scale value of each pixel of row image of the measured object image after segmentation is pressed the one-dimensional square of the pixel of corresponding line array CCD to arrangement, form a grey value profile function, be designated as F (x), then correct F (x), obtaining the grey value profile function after correcting is F ' (x).

(1) solve the overall gray level mean value of scan line image corresponding to F (x), and be designated as G

$G=\frac{\underset{x=1}{\overset{N}{\mathrm{\&Sigma;}}}F\left(x\right)}{N};$

Wherein, N is the number of pixels of current line image, now N=6144;

(2) F (x) is solved in interval sliding average, and be designated as M (x)

$M\left(x\right)=\frac{{\mathrm{\&Sigma;}}_{x-m}^{x+n}F\left(i\right)}{m+n+1};$

Wherein, m is forward slip step-length, and n slides backward step-length, total step-length step=m+n+1,0<step<N/4; I is temporary variable, now the size of step=192, m=95, n=96, step depends primarily on the numerical value of m or n, empirically chooses the best in conjunction with measured object Texture eigenvalue, excessive or too smallly all can cause by the change of adopting thing texture.

(3) the grey value profile function F after obtaining rectification is solved ' (x)

F’(x)＝G+(F(x)-M(x))。

Irregular distribution function F (x) of gray scale before rectification and irregular distribution function F ' (x) of gray scale after correcting are as shown in Figure 6, can find out that the irregular phenomenon of original gray scale has greatly improved, finally each row Image Mosaic after correcting is become the web image after a complete rectification, as shown in Figure 5.

Claims (3)

1. the quick antidote that linear array CCD camera imaging gray scale is irregular, is characterized in that: linear array CCD camera gathers measured object image, then corrects this image line by line, finally pieces together the image after rectification;

In each line period, the gray-scale value of each pixel of row image of measured object image presses the one-dimensional square of the pixel of corresponding line array CCD to arrangement, forms a grey value profile function, is designated as F (x); Then correct F (x), obtaining the grey value profile function after correcting is F ' (x), and its concrete steps of correcting are as follows:

(1) solve the overall gray level mean value of scan line image corresponding to F (x), and be designated as G

$G=\frac{\underset{x=1}{\overset{N}{\mathrm{\&Sigma;}}}F\left(x\right)}{N};$

Wherein, N is the number of pixels of current line image;

(2) F (x) is solved in interval sliding average, and be designated as M (x)

$M\left(x\right)=\frac{{\mathrm{\&Sigma;}}_{x-m}^{x+n}F\left(i\right)}{m+n+1};$

Wherein, m is forward slip step-length, and n slides backward step-length, m, total step-length step=m+n+1,0<step<N/4; I is temporary variable,

(3) the grey value profile function F after obtaining rectification is solved ' (x)

F’(x)＝G+(F(x)-M(x))。

2. the irregular quick antidote of a kind of linear array CCD camera gradation of image according to claim 1, is characterized in that, m+1=n.

3. the irregular quick antidote of a kind of linear array CCD camera gradation of image according to claim 1, is characterized in that, pixel number >=1024 of linear array CCD camera.