CN1317894C - Adaptive image frame deinterlacing device and method - Google Patents

Abstract

A adaptive image picture de-interlacing device and method comprises a line segment characteristic difference value calculating module, which calculates a line segment characteristic difference value of a line segment in an image picture; a motion vector calculation module, which calculates the motion vector of the macro block containing line segment in the image picture; an intra block number calculation module that calculates a number of intra blocks in an image picture, wherein the intra blocks represent macroblocks within which there are no motion vectors; a trigger value generation module for judging whether the number of internal blocks is greater than the first critical value so as to select the generation mode of the trigger value; and an image processing module for judging whether the trigger value is larger than the second critical value so as to select the de-interlacing algorithm to perform de-interlacing processing of the line segment.

Description

Adaptive image frame deinterlacing device and method

technical field

The invention relates to an image frame deinterlacing device and method, in particular to an adaptive image frame deinterlacing device and method.

Background technique

With the development of technology, the playback quality of audio-visual media is getting higher and higher. However, different image processing methods will affect the playback quality of audio-visual media during playback. An obvious example is the interlace scan image Progressive scan (progressive scan) image playback platform playback.

An interlaced image consists of two fields, each of which contains only the odd lines of the image or the even lines of the image, so in an interlaced image, each odd field image and even field image The resolution is only half of the original image. The advantage of the interlaced scanning method is that the dynamic image display is smooth, but the disadvantage is that the screen will flicker. In contrast, the scanning speed of progressive scanning is twice that of interlaced scanning, so the picture is relatively slender and clear. Therefore, most current audio-visual equipment scans and displays images in progressive scanning.

Today’s analog TV scanning standards include National Television System Committee (NTSC) and Phase Alternation by Line (PAL). Among them, the images of the NTSC system still use interlaced scanning. If the images of the NTSC system are directly played in progressive The scanning display system can only display the images of odd and even fields of view respectively, so the resolution of the playback image is only half of the resolution of the original image, and the playback quality of the image will be greatly reduced; on the other hand, if the interlaced scanning If the DVD image is directly played on a progressive scan display system, the aforementioned situation will also occur. In order to solve the problem that the playback quality of the interlaced scanning image is lowered when it is played on a progressive scanning playback platform, de-interlacing processing (De-interlace) must be performed on the interlaced scanning image. Deinterlacing can be said to be a method of converting an interlaced image into a progressive image.

De-interlacing includes two basic linear conversion techniques, namely Weave and Bob. As shown in FIG. 1A , the weaving process is to overlap or weave two input fields of view (an odd-numbered field of view and an even-numbered field of view) together to generate a sequential frame. As shown in Figure 1B, the swing process only accepts one of the fields of view of the input image (for example, only accepts the image of even-numbered lines), and discards the other field of view (that is, the image of odd-numbered lines), so the resolution of the picture in the vertical direction It will become half of the original image. In order to restore the image to the resolution of the original image, two adjacent scanning lines of the same field of view are used to fill in the discarded scanning lines of another field of view.

Weaving processing has a good processing effect on static images, but for dynamic images, weaving processing will produce image misalignment, so the image will appear jagged or rough, which will reduce the playback quality; on the other hand Although the wobble processing can overcome the image alignment error of the dynamic image and make the dynamic image clearer and more natural, but for the static image, the vertical resolution of the image after the wobble processing is reduced and then compensated by interpolation, making the picture Blurred easily and less sharp.

In view of this, an adaptive image frame de-interlacing method and device are provided, so that it can select an appropriate de-interlacing algorithm processing method according to the characteristics of interlaced images, so that both dynamic images or static images can be processed through better Deinterlace to improve playback quality.

Contents of the invention

In view of the above problems, the present invention provides an adaptive image frame de-interlacing method and device.

Therefore, the adaptive image deinterlacing device according to the present invention includes a LineSegment feature difference calculation module, a motion vector calculation module, an IntraBlock quantity calculation module, and a Trigger Value generation module and an image frame processing module. Wherein, the line segment feature difference calculation module calculates the line segment feature difference of the line segment in the image picture, the motion vector calculation module calculates the motion vector of the macroblock (Macro Block) that contains the line segment in the image picture, and the inner block quantity calculation module calculates the motion vector of the line segment in the image picture The number of inner blocks, wherein the inner block represents a macroblock without a motion vector in it; the trigger value generation module judges whether the number of inner blocks is greater than the first critical value, and when the number of inner blocks is greater than the first critical value, according to the line segment feature difference value to generate a trigger value, and when the number of inner blocks is not greater than the first critical value, the trigger value is generated according to the line segment feature difference and the motion vector; the image frame processing module judges whether the trigger value is greater than the second critical value, and when the trigger value is greater than When the second critical value is reached, the first de-interlacing algorithm is used to perform de-interlacing processing of the line segments, and when the trigger value is not greater than the second critical value, the second de-interlacing algorithm is used to perform de-interlacing processing of the line segments.

In addition, the adaptive image frame de-interlacing method according to the present invention includes calculating the line segment characteristic difference value of the line segment in the image frame, calculating the motion vector of the macroblock containing the line segment in the image frame, and calculating the number of inner blocks in the image frame, wherein The inner blocks represent macroblocks that have no motion vectors within them. Next, it is judged whether the number of inner blocks is greater than the first critical value. When the number of inner blocks is greater than the first critical value, a trigger value is generated according to the line segment feature difference, and when the number of inner blocks is not greater than the first critical value, the trigger value is generated according to the line segment feature difference value and motion vector to generate the trigger value. Then, it is judged whether the trigger value is greater than the second critical value. When the trigger value is greater than the second critical value, the first de-interlacing algorithm is used to perform line segment de-interlacing processing, and when the trigger value is not greater than the second critical value, the second de-interlacing algorithm is used. The de-interlacing algorithm performs de-interlacing processing of line segments.

As mentioned above, according to the adaptive image frame de-interlacing method and device of the present invention, it includes a trigger value generating module and an image frame processing module, which can target the line segment feature difference value, motion vector and inner block number in the image frame, Selecting an appropriate de-interlacing method such as a wobble algorithm or a weaving algorithm enables the image to have a better display effect.

Description of drawings

FIG. 1A and FIG. 1B are a set of schematic diagrams showing a known weaving process de-interlacing method and a wobble process de-interlacing method;

FIG. 2 is a schematic diagram showing an adaptive image frame deinterlacing device according to a preferred embodiment of the present invention;

FIG. 3 is a schematic diagram showing an adaptive image frame de-interlacing method according to a preferred embodiment of the present invention; and

FIG. 4 is a schematic diagram showing pixel values of different fields of view in a current frame and a reference frame in an adaptive image frame de-interlacing method according to a preferred embodiment of the present invention.

Description of component symbols:

11: Video image buffer module

12: Image output module

13: Image display module

2: Adaptive image de-interlacing device

21: Line segment feature difference calculation module

22:Movement vector calculation module

23: Inner block quantity calculation module

24: Trigger value generating module

25: Image screen processing module

S01-S12: Adaptive Image Deinterlacing Method

Detailed ways

An adaptive image frame deinterlacing device according to a preferred embodiment of the present invention will be described below with reference to the relevant drawings.

As shown in Figure 2, the adaptive image frame de-interlacing device 2 of the preferred embodiment of the present invention includes a segment feature difference calculation module 21, a motion vector calculation module 22, an inner block number calculation module 23, a trigger value generation module 24 and an image frame processing module 25 . In addition, the adaptive image frame deinterlacing device 2 may include a video image buffer module 11 and an image output module 12, which are respectively used to provide the image frame and output the output image generated after deinterlacing, and may also include an image display module 13, which receives and displays the output image generated by the image output module 12. Of course, other equivalent modifications or changes known to those skilled in the art should also be included in the scope of the present invention, for example, the image display module can be included in an adaptive image frame deinterlacing device (not shown in the figure).

The line segment feature difference calculation module 21 calculates the line segment feature difference of the line segment in the image picture, the motion vector calculation module 22 calculates the motion vector of the macroblock containing the line segment in the image picture, and the inner block quantity calculation module 23 calculates the inner block quantity in the image picture , where the inner blocks represent macroblocks that have no motion vectors within them. The trigger value generation module 24 judges whether the number of inner blocks is greater than the first critical value, and when the number of inner blocks is greater than the first critical value, generates a trigger value according to the line segment feature difference, and when the number of inner blocks is not greater than the first critical value, A trigger value is generated according to the feature difference of the line segment and the moving vector. The image frame processing module 25 judges whether the trigger value is greater than the second critical value, and when the trigger value is greater than the second critical value, uses the first de-interlacing algorithm to de-interlace the line segments. Wherein the first de-interlacing algorithm can be a swing (Bob) algorithm, and when the trigger value is not greater than the second critical value, utilize the second de-interlacing algorithm to carry out the de-interlacing processing of the line segment, wherein the second de-interlacing algorithm can be a weaving (Weave )algorithm.

The method for adaptive image deinterlacing according to a preferred embodiment of the present invention will be described below with reference to FIG. 3 and FIG. The description will match the aforementioned modules and their component symbols.

The adaptive image frame de-interlacing method according to a preferred embodiment of the present invention includes the following steps S01 to S12. In step S01, the image frame is read from the video image buffer module 11, wherein the image frame can be a decoded MPEG image frame, or a P frame (Predicted frame) data, and the P frame data includes a motion vector .

In step S02, after the line segment characteristic difference calculation module reads in the pixel data in the macroblock of the reference picture and the current picture, a line segment width value is selected so that in the image range corresponding to the macroblock, each scanning line can be recreated. Divided into line segments of fixed size. By comparing the line segment feature values in different pictures but at the same line segment position, and using the comparison result as the line segment feature difference value.

The line segment eigenvalues (δ _{i, n, m} ) can be calculated from the line segment of the selected size in the screen to calculate the absolute value of the pixel value difference at the relative position in the adjacent scan lines between two different fields of view, and then the absolute value of each pixel value difference After adding to produce, its calculation formula is shown in formula [1]:

${\mathrm{<span\; class="notranslate">\; \&delta;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; m</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; Dif</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; :</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; :</span>\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; j</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}<span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; Odd</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \_</span>\mathrm{<span\; class="notranslate">\; LineSeg</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \_</span>\mathrm{<span\; class="notranslate">\; Pix</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span>$

$-\mathrm{even}(\mathrm{no}+1)\_\mathrm{LineSeg}\left(m\right)\_\mathrm{Pix}\left(k\right)$ 【1】

Among them, i represents the current picture, n represents the number of scanning lines of the picture, m represents the number of line segments in this scanning line, j represents the width of the line segment (in pixels), and its width is adjustable; k represents pixels position on this line segment.

After calculating the line segment eigenvalues (δ _{i, n, m} ) of the current picture and the line segment eigenvalues (δ _{il, n, m} ) of the reference picture, the absolute value of the difference between the line segment eigenvalues of the current picture and the reference picture is is the characteristic difference of the line segment, and the calculation formula of the characteristic difference of the line segment (Δ _{i, n, m} ) is shown in formula [2]:

Δ _{i, n, m} = |(δ _{i, n, m} )-(δ _{il, nm} )| 【2】

Pixel values of different fields of view in the current frame and the reference frame. When the line segment width is 4, in the odd field of view of the current screen, the values of the 1st to 4th pixels in the first line segment are (10, 64, 70, 83), and the pixel values of the first line segment in the even field of view are The values of the 1st to 4th pixels are (13, 40, 65, 70). According to the formula [1], the feature value of the first line segment of the current screen can be calculated as 45. The calculation process is shown in the following formula [3]. In addition, in the odd-numbered field of view of the reference picture, the values of the first to fourth pixels in the first line segment are (11, 10, 64, 70), and the values of the first to fourth pixels in the first line segment in the even-numbered field of view are The pixel value is (10, 13, 40, 65), and the feature value of the first line segment of the reference picture can be calculated according to the formula [1] to be 33. The calculation process is shown in the following formula [4]. After calculating the line segment feature values of the current picture and the previous picture, calculate the line segment feature difference value of 12 according to formula [2]. The calculation process is shown in the following formula [5].

δ _{i, n, m} = ∑(|10-13|+|64-40|+70-65|+|83-70|)＝45 【3】

δ _{i-1, n, m} = ∑(|11-10|+|10-13|+|64-40|+|70-65|)＝33 【4】

Δ _{i, n, m} =|(δ _{i, n, m} )-(δ _{il, n, m} )|=|45-33|=12 【5】

After the line segment characteristic difference is calculated in step S02, in step S03, the motion vector calculation module reads the motion vector in the macroblock, which is located in the same macroblock as the line segment processed by the line segment feature difference calculation module. Since the motion vector value is to record the displacement of a specific block between different fields of view at the same time, the motion vector is represented by a specific coordinate value (coordinates value), and the movement of the block is read sequentially Vector value, the picture can be reconstructed. In addition, the block size can generally be 8×8 pixels or 16×16 pixels, and pixels of other sizes can also be used to form blocks, which are adjusted depending on the performance of the device performing de-interlacing processing and the size of the memory provided.

After the motion vector is captured, the motion vector calculation module adds the length of the X axis of the motion vector to the length of the Y axis of the motion vector to obtain a rectangular length of the motion vector. For example, the extracted motion vector value is (1, -32), which means that the block is displaced by 1 pixel in the positive direction of the X axis, and 32 pixels in the negative direction of the Y axis direction. The length of the right angle of this motion vector is 33 , its calculation is as formula [6]:

|1|+|-32|＝33【6】

After calculating the right-angle length of the motion vector in step S03, go to step S04. In step S04, the calculation module for the number of inner blocks calculates the number of all macroblocks marked as inner blocks in a complete image frame. If the macroblock is marked as an inner block, it means that there is no motion vector in the macroblock, so the image content of the macroblock is not composed of the motion vector and the block referenced by the motion vector, but directly composed of the pixel data recorded in the macroblock , in other words, the more macroblocks in the image frame marked as intra blocks, the less correlation the image frame has with the reference frame.

In step S05, the trigger value generation module determines whether the number of inner blocks calculated by the inner block number calculation module in step S04 is greater than the first critical value. If the number of inner blocks is greater than the first critical value, then enter step S06, and in step S06, generate a trigger value according to the line segment characteristic difference; if the inner block number is not greater than the first critical value, then enter step S07, and In step S07, a trigger value is generated according to the feature difference of the line segment and the motion vector. After executing step S06 or step S07 to generate a trigger value, go to step S08.

In step S08, the image frame processing module determines whether the trigger value generated in step S06 or step S07 is greater than a second critical value. If the trigger value is greater than the second critical value, then enter step S09, and in step S09, for the current picture line segment processed in step S02, use the first de-interlacing algorithm to perform de-interlacing processing, wherein the first de-interlacing algorithm can be It is a swing (Bob) algorithm; if the trigger value is not greater than the second critical value, then enter step S10, and in step S10, utilize the second de-interlacing algorithm to perform de-interlacing for the current picture line segment processed in step S02 Processing, wherein the second de-interleaving algorithm may be a Weave algorithm. After executing step S09 or step S10 to execute the de-interleaving algorithm, go to step S11.

In step S11, it is judged whether the entire image frame has been processed, if there are unprocessed line segments and images, proceed to step S02 to continue processing, and if the process has been completed, proceed to step S12.

In step S12, the de-interlacing processing flow ends, and the de-interlaced image can be output to the image output module, so that the de-interlaced image can be formed into an output image, and then output to various display devices.

In summary, according to the method and device for adaptive image frame de-interlacing of the present invention, it includes a trigger value generation module and an image frame processing module, which can target the line segment feature difference value, motion vector and inner block number in the image frame, Selecting an appropriate de-interlacing method such as a wobble algorithm or a weaving algorithm enables the image to have a better display effect.

The above description is for illustration only, not for limitation. Any equivalent modification or change without departing from the spirit and scope of the present invention shall be included in the appended claims.

Claims (13)

1, a kind of adaptive image frame deinterlace method comprises:

Calculate the line segment feature difference of a line segment in the image frame;

Calculate the motion-vector that comprises the macro block of described line segment in the described image frame;

Calculate the interior number of blocks in the described image frame, wherein said interior piece represents not have in it macro block of motion-vector;

Judge that whether described interior number of blocks is greater than one first critical value;

When number of blocks is greater than described first critical value in described, produce a trigger value according to described line segment feature difference;

When number of blocks is not more than described first critical value in described, produce described trigger value according to described line segment feature difference and described motion-vector;

Judge that whether described trigger value is greater than one second critical value;

When described trigger value during greater than described second critical value, the deinterleave that utilizes first deinterlace algorithm to carry out described line segment is handled; And

When described trigger value was not more than described second critical value, the deinterleave that utilizes second deinterlace algorithm to carry out described line segment was handled.

2, adaptive image frame deinterlace method as claimed in claim 1, wherein said line segment feature difference are to deduct a corresponding reference picture line segment feature value by a present line segment feature value to be calculated.

3, adaptive image frame deinterlace method as claimed in claim 1, wherein said motion-vector comprises the angle length of described motion-vector, and the angle length of described motion-vector is the Y-axis length that the X-axis length of described motion-vector adds the above motion-vector.

4, adaptive image frame deinterlace method as claimed in claim 1, the quantity of piece in all macro blocks that wherein said interior number of blocks is described image frame are marked as.

5, adaptive image frame deinterlace method as claimed in claim 1, wherein said image frame are a mpeg image picture or are a P frame data.

6, adaptive image frame deinterlace method as claimed in claim 1, wherein said first deinterlace algorithm are a swing algorithm, and described second deinterlace algorithm is a braiding algorithm.

7, adaptive image frame deinterlace method as claimed in claim 1 more comprises:

Handle formation one output image according to the deinterleave of described line segment.

8, a kind of adaptive image frame deinterlace device comprises:

One line segment feature difference calculating module, a line segment feature difference of its computed image picture middle conductor;

One motion-vector computing module, it calculates a motion-vector of the macro block that comprises described line segment in the described image frame;

Number of blocks computing module in one, it calculates the interior number of blocks in the described image frame, and wherein said interior piece represents not have in it macro block of motion-vector;

One trigger value generation module, it judges that whether described interior number of blocks is greater than one first critical value, and when number of blocks is greater than described first critical value in described, produce a trigger value according to described line segment feature difference, and when number of blocks is not more than described first critical value in described, produce described trigger value according to described line segment feature difference and described motion-vector; And

One image frame processing module, it judges that whether described trigger value is greater than one second critical value, and when described trigger value during greater than described second critical value, the deinterleave that utilizes first deinterlace algorithm to carry out described line segment is handled, and when described trigger value was not more than described second critical value, the deinterleave that utilizes second deinterlace algorithm to carry out described line segment was handled.

9, adaptive image frame deinterlace device as claimed in claim 8, wherein said line segment feature difference calculating module are to deduct a corresponding reference picture line segment feature value with a present line segment feature value to calculate described line segment feature difference.

10, adaptive image frame deinterlace device as claimed in claim 8, wherein said motion-vector computing module are the degree of angular length always with the described motion-vector of Y-axis length computation of the X-axis length of described motion-vector and described motion-vector.

11, adaptive image frame deinterlace device as claimed in claim 8, the quantity of piece in all macro blocks that wherein said interior number of blocks computing module calculates described image frame are marked as.

12, adaptive image frame deinterlace device as claimed in claim 8, wherein said image frame are a mpeg image picture or are a P frame data.

13, adaptive image frame deinterlace device as claimed in claim 8 more comprises:

One image output module, its deinterleave according to described line segment is handled to constitute an output image.

Images