CN103024328B - A kind of method improving screenshot quality of digital video recorder - Google Patents

Abstract

The present invention discloses and is not a kind ofly increasing the grabgraf quality improving image under hardware cost, be applied to the method for the raising screenshot quality of digital video recorder of monitor service better, a kind of brightness/gray scale subregion linear expansion algorithm for image enhancement based on weight is applied in digital video recorder DVR grabgraf processing method, before generation grabgraf file, by a kind of brightness/gray scale subregion linear expansion algorithm for image enhancement based on weight, image enhaucament computing is carried out to improve picture quality to raw image data, again the view data after image enhaucament computing is stored as image file afterwards, wherein digital video recorder DVR grabgraf processing method comprises: obtain raw image data, brightness/gray scale statistics of extremes, input image lightness gray threshold judges, input image lightness gray scale subregion, input picture subregion brightness/gray scale statistics, output image brightness/gray scale subregion, based on the brightness/gray scale subregion linear expansion of weight, view data after stores processor.

Description

A kind of method improving screenshot quality of digital video recorder

Technical field

The invention belongs to the technical field of video monitoring, be specifically related to a kind of method improving screenshot quality of digital video recorder.

Background technology

DVR (DigitalVideoRecorder), i.e. digital video recorder, it is a set of computer system carrying out compressing to voice data video data stores processor, have and voice messaging and image information are recorded for a long time, recorded a video, the functions such as retrieval video and audio recording information, grabgraf, telemonitoring and control ancillary equipment.

DSP (DigitalSignalProcessor), i.e. digital signal processor is a core devices in DVR.The functions such as the catching of audio-video signal in DVR equipment, the compressed encoding of audio-video signal, the decoding of audio-video signal have been responsible for by this device.

Grabgraf functional module is the necessary functions module of DVR equipment, the function of grabgraf module is realized by DSP, specific implementation process is, DSP receives image grabbing command, DSP captures the image of expection according to image grabbing command, DSP sends the image data grabbed to DVR master controller, and the pictorial information received is saved as the picture that computer can identify by DVR master controller.

In DVR, the general operation method of grabgraf module is: DVR master controller sends image grabbing command, DSP catches the original video data of the yuv format of current time after receiving the image grabbing command of master controller transmission, yuv format is stored as bmp form through color space conversion or is stored as jpg form through compression algorithm process by DSP.

When partially bright, partially dark, the vaporific situation such as fuzzy appears in the content in the video scene needing grabgraf, the effect of grabgraf is just not ideal, and user is difficult to the valuable information got in video, does not reach the grabgraf effect of expection.

Summary of the invention

Technology of the present invention is dealt with problems and is: overcome the deficiencies in the prior art, provides a kind of and can improve the grabgraf quality of image when not increasing hardware cost, is applied to the method for the raising screenshot quality of digital video recorder of monitor service better.

Technical solution of the present invention is: the method for this raising screenshot quality of digital video recorder, a kind of brightness/gray scale subregion linear expansion algorithm for image enhancement based on weight is applied in digital video recorder DVR grabgraf processing method, before generation grabgraf file, by a kind of brightness/gray scale subregion linear expansion algorithm for image enhancement based on weight, image enhaucament computing is carried out to improve picture quality to raw image data, again the view data after image enhaucament computing is stored as image file afterwards;

Wherein digital video recorder DVR grabgraf processing method comprises the following steps:

(1) raw image data is obtained;

(2) brightness/gray scale statistics of extremes;

(3) input image lightness gray threshold judges;

(4) input image lightness gray scale subregion;

(5) input picture subregion brightness/gray scale statistics;

(6) output image brightness/gray scale subregion;

(7) based on the brightness/gray scale subregion linear expansion of weight;

(8) view data after stores processor.

This method is applied on DVR equipment, the problem that grabgraf effect is undesirable when there is partially bright, partially dark, the vaporific situation such as fuzzy in the video scene needing grabgraf can be solved, when not increasing hardware cost, better picture quality effect can be obtained, improve the grabgraf quality of image.

Accompanying drawing explanation

Fig. 1 is the general processing method sketch of grabgraf in DVR equipment;

Fig. 2 is the processing method sketch of grabgraf after DVR equipment application the inventive method;

Fig. 3 is the detail flowchart of the processing method of grabgraf after DVR equipment application the inventive method.

Embodiment

The method of this raising screenshot quality of digital video recorder, a kind of brightness/gray scale subregion linear expansion algorithm for image enhancement based on weight is applied in digital video recorder DVR grabgraf processing method, before generation grabgraf file, by a kind of brightness/gray scale subregion linear expansion algorithm for image enhancement based on weight, image enhaucament computing is carried out to improve picture quality to raw image data, again the view data after image enhaucament computing is stored as image file afterwards;

Wherein digital video recorder DVR grabgraf processing method comprises the following steps:

(1) raw image data is obtained;

(2) brightness/gray scale statistics of extremes;

(3) input image lightness gray threshold judges;

(4) input image lightness gray scale subregion;

(5) input picture subregion brightness/gray scale statistics;

(6) output image brightness/gray scale subregion;

(7) based on the brightness/gray scale subregion linear expansion of weight;

(8) view data after stores processor.

This method is applied on DVR equipment, the problem that grabgraf effect is undesirable when there is partially bright, partially dark, the vaporific situation such as fuzzy in the video scene needing grabgraf can be solved, when not increasing hardware cost, better picture quality effect can be obtained, improve the grabgraf quality of image.

Preferably, in step (1): raw image data adopts yuv format.

Preferably, in step (2): the distribution of the brightness/gray scale of all pixels of original image is added up, draw minimum value and the maximum of original image brightness/gray scale, namely travel through the brightness/gray scale of original image and compare size, obtaining brightness/gray scale minimum value nYPixelMin and the brightness/gray scale maximum nYPixelMax of original image.

Preferably, in step (3): according to brightness/gray scale minimum value nYPixelMin and brightness/gray scale maximum nYPixelMax, and the original image preset needs brightness/gray scale minimum threshold nYUserMin to be processed and brightness/gray scale max-thresholds nYUserMax, determine the brightness/gray scale interval (nYSrcMin, nYSrcMax) needing to carry out region linear expansion.

Preferably, in step (4): will the brightness/gray scale interval (nYSrcMin needing to carry out linear expansion be in, nYSrcMax) gray value in is divided into the identical nBlock of length interval, the value of nBlock is set by the user, the starting point nYSrcGray of each original image brightness/gray scale subregion _{i, 0}represent, the terminal nYSrcGray of each original image brightness/gray scale subregion _i.1represent.

Preferably, in step (5): in the brightness/gray scale subregion section determined, add up to the number of pixels that brightness/gray scale is distributed in each section respectively, corresponding each brightness/gray scale subregion, obtains being distributed in the number of pixels nYSrcBlockPixel in this brightness/gray scale subregion respectively _i.By all nYSrcBlockPixel _ibe added, obtain the pixel count sum nYSrcBlockPixelTotal be distributed in all brightness/gray scale subregions.

Preferably, in step (6): the gray value be in target image brightness/gray scale interval (nYDstMin, nYDstMax) is divided into the different nBlock of length interval, each length of an interval degree YDstBlockLength _iby weights W _ibe multiplied with output image brightness/gray scale total length and obtain, weights W _iby the number of pixels nYSrcBlockPixel in each subregion in input picture _iobtain divided by the sum of all pixels nYSrcBlockPixelTotal be distributed in all brightness/gray scale subregions of input picture, the starting point nYDstGray of each target image brightness/gray scale subregion _{i, 0}represent, the terminal nYDstGray of each target image brightness/gray scale subregion _i.1represent.

Preferably, in step (7): set up mapping relations between input image lightness gray scale subregion and output image brightness subregion, that is, first calculate the mapping ratio α of each brightness/gray scale subregion according to formula (1) _i

${\mathrm{\α}}_i=\frac{{\mathrm{nYDstGray}}_{i,1}-n{\mathrm{YDstGray}}_{i,0}}{{\mathrm{nYSrcGray}}_{i,1}-{\mathrm{nYSrcGray}}_{i,0}}---\left(1\right)$

Then, carrying out mapping for needing luminance grayscale values to be processed in original image and export, directly exporting for not needing luminance grayscale values to be processed in original image.

Preferably, in step (8): the view data of the yuv format after process is stored as bmp form through color space conversion or is stored as jpg form through overcompression.

Be described with regard to a preferred embodiment of the present invention now.

In DVR equipment, the general process flow of grabgraf as shown in Figure 1.General processing method comprises raw image data and obtains and view data storage.After the raw image data being got monitoring scene by the DSP module in DVR equipment, data store according to the setting of system, are stored as the target image of jpg form after the raw image data of the yuv format got is stored as the target image of bmp form or compressed algorithm process after color space conversion.

Fig. 2 is the process flow sketch of grabgraf after DVR equipment application the inventive method.Image enhaucament is applied in the grabgraf of DVR equipment by the present invention, effectively can improve the picture quality of grabgraf picture, effectively can solve the problem that grabgraf effect is undesirable when there is partially bright, partially dark, the vaporific situation such as fuzzy in the video scene needing grabgraf.

Fig. 3 is the processing method detail flowchart of grabgraf after DVR equipment application the inventive method, see Fig. 3, is introduced, comprises the following steps the embodiment of the application:

(1) pass through the raw image data of image data acquisition to yuv format of DSP.

(2) add up the distribution of the brightness/gray scale of whole original image, draw minimum value and the maximum of image brightness gray scale, concrete grammar is:

Respectively the brightness/gray scale original maximum nYPixelMax of the pixel of image is set to 0, the original intensity minimum gray value nYPixelMin of image pixel is set to 255.Afterwards entire image is traveled through, obtain the luminance grayscale values nYVal of each pixel.If the luminance grayscale values nYVal of pixel is greater than nYPixelMax, then by the luminance grayscale values nYVal assignment of this point to nYPixelMax, if the luminance grayscale values nYVal of pixel is less than or equal to nYPixelMax, then the value of nYPixelMax remains unchanged; If the luminance grayscale values nYVal of pixel is less than nYPixelMin, then by the luminance grayscale values nYVal assignment of this point to nYPixelMin, if the luminance grayscale values of pixel is greater than or equal to nYPixelMin, then the value of nYPixelMin remains unchanged.

After having traveled through this width image, brightness/gray scale minimum value nYPixelMin and the brightness/gray scale maximum nYPixelMax of entire image can be obtained.

(3) determine that the original image brightness/gray scale needing to carry out brightness/gray scale linear expansion is interval.After calculating original image brightness/gray scale extreme value (brightness/gray scale minimum value nYPixelMin and brightness/gray scale maximum nYPixelMax), original image in conjunction with user's setting needs brightness/gray scale threshold value extreme value to be processed (brightness/gray scale minimum threshold nYUserMin and brightness/gray scale max-thresholds nYUserMax), determine the brightness/gray scale interval (nYSrcMin, nYSrcMax) needing to carry out region linear expansion.In the preferred scheme of one, the original image of user's setting needs brightness/gray scale threshold value extreme value interval to be processed to be (15,240).Concrete grammar is:

$\mathrm{nYSrcMin}=\left\{\begin{array}{cc}\mathrm{nYPixelMin}& (\mathrm{nYPixelMin}>\mathrm{nYUserMin})\\ \mathrm{nYUserMin}& (\mathrm{nYPixelMin}<\mathrm{nYUserMIn})\end{array}\right.$

$\mathrm{nYSrcMax}=\left\{\begin{array}{cc}\mathrm{nYPixelMax}& (\mathrm{nYPixelMax}<\mathrm{nYUserMax})\\ \mathrm{nYUserMax}& (\mathrm{nYpixelMax}>\mathrm{nYUserMax})\end{array}\right.$

(4), needing the gray value carried out in the brightness/gray scale interval (nYSrcMin, nYSrcMax) of linear expansion to be divided into nBlock interval, the value of nBlock is set by the user, and in the preferred scheme of one, the value of nBlock is 8.Concrete grammar is:

Calculate input picture partition length:

YSrcBlockLength=(nYSrcMax-nYSrcMin)/nBlock

The starting point nYSrcGray of first brightness/gray scale subregion _1,0with terminal nYSrcGray _1.1be shown below:

$\left\{\begin{array}{c}{\mathrm{nYSrcGray}}_{\mathrm{1,0}}=\mathrm{nYSrcMin}\\ {\mathrm{nYSrcGray}}_{\mathrm{1,1}}=\mathrm{nYSrcMin}+\mathrm{YSrcBlockLength}\end{array}\right.$

The starting point nYSrcGray of i-th brightness/gray scale subregion _{i, 0}with terminal nYSrcGray _i.1be shown below:

$\left\{\begin{array}{cc}{\mathrm{nYSrcGray}}_{i,0}={\mathrm{nYSrcGray}}_{i-\mathrm{1,1}}& (1<i<\mathrm{nBlock})\\ {\mathrm{nYSrcGray}}_{\mathrm{1,1}}={\mathrm{nYSrcGray}}_{i,0}+\mathrm{YSrcBlockLength}& (1<i<\mathrm{nBlock})\end{array}\right.$

The place nYSrcGray of the n-th Block brightness/gray scale subregion _{nBlock, 0}with terminal nYSrcGray _nBlock.1be shown below:

$\left\{\begin{array}{c}{\mathrm{nYSrcGray}}_{\mathrm{nBlock},0}={\mathrm{nYSrcGray}}_{\mathrm{nBlock}-1,1}\\ {\mathrm{nYSrcGray}}_{\mathrm{nBlock},1}=\mathrm{nYSrcMax}\end{array}\right.$

(5) respectively the number of pixels in each brightness/gray scale section is added up, obtain respectively being distributed in the number of pixels nYSrcBlockPixel in each brightness/gray scale subregion _i, by all nYSrcBlockPixel _ibe added, obtain the pixel count sum nYSrcBlockPixelTotal be distributed in all brightness/gray scale subregions, concrete grammar is:

$\mathrm{nYSrcBlockPixelTotal}=\underset{i=1}{\overset{\mathrm{nBlock}}{\mathrm{\&Sigma;}}}{\mathrm{nYSrcBlockPixel}}_i$

(6) the gray value be in target image brightness/gray scale interval (nYDstMin, nYDstMax) is divided into the different nBlock of length interval.In the preferred scheme of one, the value of nBlock is 8, is (0,220) between object brightness gray area.The length YDstBlockLength of each subregion _iby weights W _ibe multiplied with output image brightness/gray scale total length and obtain.Concrete grammar is:

YDstBlockLength _i=W _i×(nYDstMax-nYDstMin)

Weights W _iby the number of pixels nYSrcBlockPixel in each subregion in input picture _iobtain divided by the sum of all pixels nYSrcBlockPixelTotal be distributed in all brightness/gray scale subregions of input picture, circular is shown below:

$W_i=\frac{\mathrm{nYSrcBlock}{\mathrm{Pixel}}_i}{\mathrm{nYSrcBlockPixelTotal}}$

The starting point nYDstGray of first brightness/gray scale subregion _1,0with terminal nYDstGray _1.1be shown below:

$\left\{\begin{array}{c}{\mathrm{nYDstGray}}_{\mathrm{1,0}}=\mathrm{nYDstMin}\\ {\mathrm{nYDstGray}}_{\mathrm{1,1}}=\mathrm{nYSrcMin}+{\mathrm{YDStBlockLength}}_1\end{array}\right.$

The starting point nYDstGray of i-th brightness/gray scale subregion _{i, 0}with terminal nYDstGray _i.1be shown below:

$\left\{\begin{array}{cc}{\mathrm{nYDstGray}}_{i,0}={\mathrm{nYDstGray}}_{i-\mathrm{1,1}}& (1<i<\mathrm{nBlock})\\ {\mathrm{nYDstGray}}_{\mathrm{1,1}}={\mathrm{nYDstGray}}_{i,0}+{\mathrm{YDstBlockLength}}_i& (1<i<\mathrm{nBlock})\end{array}\right.$

The place nYDstGrat of the n-th Block brightness/gray scale subregion _{nBlock, 0}with terminal nYDstGray _nBlock.1be shown below:

$\left\{\begin{array}{c}{\mathrm{nYDstGray}}_{\mathrm{nBlock},0}={\mathrm{nYDstGray}}_{\mathrm{nBlock}-1,1}\\ {\mathrm{nYDstGray}}_{\mathrm{nBlock},1}=\mathrm{nYDstMax}\end{array}\right.$

(7) between input image lightness gray scale subregion and output image brightness subregion, set up mapping relations.First calculate the mapping ratio α of each brightness/gray scale subregion _i:

${\mathrm{\&alpha;}}_i=\frac{{\mathrm{nYDstGray}}_{i,1}-{\mathrm{nYDstGray}}_{i,0}}{{\mathrm{nYSrcGray}}_{i,1}-{\mathrm{nYSrcGray}}_{i,0}}$

Luminance grayscale values in original image is in the luminance grayscale values nYSrc of the pixel in (nYSrcMin, nYSrcMax) region _k, first judge nYSrc _kbe in which subregion of input image lightness gray scale subregion, suppose after judging, gray value nYSrc _kbe in i-th brightness/gray scale subregion, be then handled as follows:

nYDst _k=nYDstGray _i,0+α _i×(nYSrc _k-nYDstGray _i,0)

Luminance grayscale values in original image is in the luminance grayscale values nYSrc of (nYSrcMin, nYSrcMax) extra-regional pixel _m, then do not process, directly export.

nYDst _m=nYSrc _m

(8) the view data of yuv format is after treatment stored as bmp form through color space conversion or is stored as jpg form through overcompression.

The above; it is only preferred embodiment of the present invention; not any pro forma restriction is done to the present invention, every above embodiment is done according to technical spirit of the present invention any simple modification, equivalent variations and modification, all still belong to the protection range of technical solution of the present invention.

Claims (2)

1. one kind is improved the method for screenshot quality of digital video recorder, it is characterized in that, a kind of brightness/gray scale subregion linear expansion algorithm for image enhancement based on weight is applied in digital video recorder DVR grabgraf processing method, before generation grabgraf file, by a kind of brightness/gray scale subregion linear expansion algorithm for image enhancement based on weight, image enhaucament computing is carried out to improve picture quality to raw image data, again the view data after image enhaucament computing is stored as image file afterwards;

Wherein digital video recorder DVR grabgraf processing method comprises the following steps:

(1) obtain raw image data, raw image data adopts yuv format;

(2) brightness/gray scale statistics of extremes: the distribution of the brightness/gray scale of all pixels of original image is added up, draw minimum value and the maximum of original image brightness/gray scale, namely travel through the brightness/gray scale of original image and compare size, obtaining brightness/gray scale minimum value nYPixelMin and the brightness/gray scale maximum nYPixelMax of original image;

(3) input image lightness gray threshold judges: according to brightness/gray scale minimum value nYPixelMin and brightness/gray scale maximum nYPixelMax, and the original image preset needs brightness/gray scale minimum threshold nYUserMin to be processed and brightness/gray scale max-thresholds nYUserMax, determine the brightness/gray scale interval (nYSrcMin, nYSrcMax) needing to carry out region linear expansion;

(4) input image lightness gray scale subregion: will the brightness/gray scale interval (nYSrcMin needing to carry out linear expansion be in, nYSrcMax) gray value in is divided into the identical nBlock of length interval, the value of nBlock is set by the user, the starting point nYSrcGray of each original image brightness/gray scale subregion _{i, 0}represent, the terminal nYSrcGray of each original image brightness/gray scale subregion _{i, 1}represent;

(5) input picture subregion brightness/gray scale statistics: in the brightness/gray scale subregion section determined, respectively the number of pixels that brightness/gray scale is distributed in each section is added up, corresponding each brightness/gray scale subregion, obtains being distributed in the number of pixels nYSrcBlockPixel in this brightness/gray scale subregion respectively _i; By all nYSrcBlockPixel _ibe added, obtain the pixel count sum nYSrcBlockPixelTotal be distributed in all brightness/gray scale subregions;

(6) output image brightness/gray scale subregion: the gray value be in target image brightness/gray scale interval (nYDstMin, nYDstMax) is divided into the different nBlock of length interval, each length of an interval degree YDstBlockLength _iby weights W _ibe multiplied with output image brightness/gray scale total length and obtain, weights W _iby the number of pixels nYSrcBlockPixel in each subregion in input picture _iobtain divided by the sum of all pixels nYSrcBlockPixelTotal be distributed in all brightness/gray scale subregions of input picture, the starting point nYDstGray of each target image brightness/gray scale subregion _{i, 0}represent, the terminal nYDstGray of each target image brightness/gray scale subregion _{i, 1}represent;

(7) based on the brightness/gray scale subregion linear expansion of weight: set up mapping relations between input image lightness gray scale subregion and output image brightness subregion, that is, the mapping ratio α of each brightness/gray scale subregion is first calculated according to formula (1) _i

${\mathrm{\&alpha;}}_i=\frac{{\mathrm{nYDstGray}}_{i,1}-{\mathrm{nYDstGray}}_{i,0}}{{\mathrm{nYSrcGray}}_{i,1}-{\mathrm{nYSrcGray}}_{r,0}}---\left(1\right)$

Then, carrying out mapping for needing luminance grayscale values to be processed in original image and export, directly exporting for not needing luminance grayscale values to be processed in original image;

(8) view data after stores processor.

2. the method for raising screenshot quality of digital video recorder according to claim 1, it is characterized in that, in step (8): the view data of the yuv format after process is stored as bmp form through color space conversion or is stored as jpg form through overcompression.