CN1662035A - Film mode detection device and related method for displacement vector estimation - Google Patents

Abstract

The invention provides a device and a method for judging a film mode by utilizing displacement estimation, which mainly adopt a displacement estimation mode to estimate the displacement and change of each target pixel in another field (such as a previous field or a next field) by taking each pixel as a target pixel in each field of video data so as to judge whether the field has the displacement, and judge whether the video data is in a film mode according to whether the field has the displacement or not.

Description

Film mode detection apparatus and related method for displacement vector estimation

technical field

The present invention relates to a device and method for film mode detection, in particular to a device and method for judging a film mode by using motion estimation.

Background technique

In traditional interlaced scanning (interlaced scanning) technology, when each image frame of a movie is to be displayed, the two image fields that make up the frame are actually interlaced. It stays put, so you won't feel any abnormality. Of the two fields, one is an odd field composed of odd scan lines, and the other is an even field composed of even scan lines. The recent progressive scanning technology, also known as non-interlaced scanning, combines two fields belonging to the same frame into one frame, and then doubles the horizontal scanning frequency. The frames are scanned sequentially, so that the quality and stability of the picture will be greatly improved.

Correctly detect the relationship between each image field of the video data, and judge whether the video data is a video mode (for example, it is a 3-2 pull down video mode, or a 2-2 pull down movie mode) and apply proper de-interlacing (such as merging two fields), such processing is necessary.

Please refer to US Patent No. 4982280 and No. 6580463 for related known technologies.

Contents of the invention

Therefore, the object of the present invention is to provide a device and method for detecting a film mode by using displacement estimation to determine a film mode.

The present invention provides a method and related device for judging a film mode by using displacement estimation. The method includes: performing at least one displacement estimation and judging step, so as to receive two image fields in the video data as one respectively. A first image field and a second image field; performing a step of target pixel displacement estimation and judgment, to divide at least one target pixel area in the first image field, and perform displacement estimation and determination for each target pixel area Judging, to provide a first judgment value corresponding to the target pixel area according to the result of the displacement estimation; and then performing a field displacement judgment step, to generate a corresponding to the first judgment value according to each target pixel area A second judgment value of the first field. After performing each displacement estimation and judging steps, according to the corresponding second judging values obtained by taking different fields in the video data as input first fields respectively, it is judged whether the video data is a film mode, And which map should be merged with the first field. The film mode detection device includes: at least one displacement estimation and judgment module, the displacement estimation and judgment module includes: a pixel displacement estimation unit, which receives a first image field and a second image in the video data field, and provide a displacement vector corresponding to the target pixel area according to a target pixel area of the first image field and a plurality of pixel areas in the second image field; a field displacement judgment unit coupled to the pixel A displacement estimation unit generates a second judgment value corresponding to the first field according to a corresponding displacement vector in the first field; and a film mode detector coupled to the at least one displacement estimation and the judging module receives the at least one second judging value and generates a judging result.

The present invention can more accurately determine the video mode to which a video data belongs, and can further select a correct de-interlacing method to correctly process the video data.

Description of drawings

FIG. 1 is a functional block diagram of an embodiment of the film mode detection device of the present invention.

FIG. 2 , FIG. 3 , FIG. 4 and FIG. 5 are flow charts of the film mode detection device in FIG. 1 performing film mode detection according to the technical spirit of the present invention.

Description of reference symbols

200: Film mode detection device 10, 20, 30: Displacement estimation and judgment module

40: Movie mode detector 11, 21, 31: Pixel displacement estimation unit

12, 22, 32: Image field displacement judgment unit 13, 23, 33: Pixel displacement estimator

14, 16, 24, 26, 34, 36: Judgment 15, 25, 35: Image field displacement estimator

17, 27, 37: Counter 61, 62, 63, 64, 65, 66: Threshold

Detailed ways

Please refer to FIG. 1 . FIG. 1 is a functional block diagram of an embodiment of a video mode detection device 200 of the present invention. Generally speaking, interlaced scanning video data is formed by concatenating various video fields, and the video mode detection device 200 of the present invention is a video that uses motion estimation between video fields to determine the input video data. mode (film mode). As shown in FIG. 1 , the film mode detection device 200 includes three displacement estimation and judgment modules 10 , 20 and 30 , and a film mode detector 40 . Wherein, the structures of the displacement estimation and judging modules 10 , 20 and 30 are substantially the same. Take the displacement estimation and determination module 10 as an example: the displacement estimation and determination module 10 includes a pixel displacement estimation unit 11 and a field displacement determination unit 12; the input end of the pixel displacement estimation unit 11 can receive two images The fields are a first image field and a second image field, and a plurality of target pixel areas are divided in the first image field, so as to perform displacement estimation and judgment for each target pixel area of the first image field, and estimate Displacement and change of the target pixel area in the second image field, and provide a first judgment value D ₁ corresponding to the target pixel area according to the result of displacement estimation. In a preferred embodiment of the present invention, the target pixel area may contain only one pixel, so it may be referred to as the target pixel for short.

In order to realize the function of the pixel displacement estimation unit 11 , a pixel displacement estimation unit 13 and a first determination unit 14 are provided therein. The pixel displacement estimator 13 is used to perform displacement estimation for each target pixel in the first image field, and estimate that the target pixel will fall in a certain position in the second image field, and obtain a corresponding displacement prediction Estimate the result. The displacement prediction result may include a motion vector MV_1 and a similarity S ₁ . The first determiner 14 is coupled to the pixel displacement estimator 13 . The first determiner 14 can be implemented with a first threshold (threshold) 61, which is used to generate the magnitude relationship between the displacement vector MV_1 and the similarity _S1 output by the pixel displacement estimator 13 and a first threshold. A first judgment value D ₁ corresponding to the target pixel is used to represent whether the target pixel has been displaced. If the value of the displacement vector MV_1 is 0 and the similarity S ₁ is greater than the first critical value, then It means that the target pixel has no displacement, and the value of the first judgment value _D1 is 0; otherwise, it means that the target pixel has displacement, and the value of the first judgment value _D1 should be set to 1.

As for the field displacement judging unit 12, it is coupled to the pixel displacement estimation unit 11, which can synthesize the first judgment values _D1 corresponding to all the target pixels in the first field to generate the corresponding first judgment value D1. A second judgment value D ₂ of the field. In order to realize the above functions, the field shift judging unit 12 may be provided with a one field information device 15 and a second judging unit 16 . The field displacement estimator 15 includes a first counter 17, which is used to accumulatively generate the corresponding first judgment value _D1 corresponding to each target pixel in the first field. A first count value C ₁ ; and the first count value C ₁ represents the number of displaced target pixels in the first field. For example, if the first judgment value _D1 of a target pixel in the first field is 1, then add ₁ to the first count value C1; if the first judgment value _D1 is 0, it means that the There is no displacement of the target pixel, so the value of the first count value _C1 remains unchanged. According to this operating principle, the first count value C ₁ can be accumulated sequentially for each target pixel in the first field.

The second judging unit 16 is coupled to the field displacement estimator 15, and is used to generate a second judgment corresponding to the first field according to the first count value C ₁ output by the field displacement estimator 15 Value D ₂ . The second judging unit 16 can be implemented with a second threshold 62 for comparing the first count value _C1 with a second critical value to generate a second judging value _D2 corresponding to the first field . The second judgment value _D2 represents whether there is field displacement in the first field, and when the value of the second judgment value _D2 is greater than the second critical value, it means that there is field displacement in the first field, so The value of the second judgment value D ₂ is set to 1, otherwise, it means that there is no field shift in the first field, so the value of the second judgment value D ₂ can be set to 0.

All devices and functions included in the displacement estimation and judgment modules 20 and 30 are basically similar to all corresponding devices and functions in the displacement estimation and judgment module 10 . In general, the video mode detection device 200 receives three adjacent fields in the video data as three front, middle and back fields: F _n-1 , F _n and F _n+1 . As shown in Fig. 1, the displacement estimation and judging module 10 uses the first image field and the second image field as inputs of the image fields _Fn and Fn _-1 respectively, and targets each target pixel in the image field _Fn A corresponding first judgment value D ₁ is generated, and a second judgment value D ₂ is generated according to each first judgment value D ₁ . Similarly, the displacement estimation and judging module 20 takes the first image field and the second image field respectively as inputs of the image fields _Fn and Fn ₊₁ , so as to generate corresponding pixel values for each target pixel in the image field _Fn . The third judgment value D ₃ , and a fourth judgment value D ₄ is generated according to each third judgment value D ₃ . As for the displacement estimation and judgment module 30, the first image field and the second image field are respectively input with the image fields Fn _-1 and Fn ₊₁ , so as to target each target pixel in the image field Fn _-1 A corresponding fifth judgment value D ₅ is generated, and a sixth judgment value D ₆ is obtained by accumulating and comparing each fifth judgment value D ₅ .

The film mode detector 40, coupled to the displacement estimation and judgment modules 10, 20 and 30, is used to generate a seventh judgment value according to the second, fourth and sixth judgment values D ₂ , D ₄ and D ₆ D ₇ and an eighth judgment value D ₈ . The seventh judgment value _D7 represents whether the video data to which the field _Fn belongs is a video mode, for example, it is a 3-2 pull down video mode or a 2-2 pull down video mode. The eighth judgment value represents whether the field _Fn should be merged with the field Fn _-1 or the field Fn ₊₁ .

The above is a preferred embodiment of the present invention. In addition, in another implementation state of the present invention, only one displacement estimation and judgment module 10 can be used to cooperate with the film mode detector 40, and the input is the first image field F _n and the second image field Fn _-1 , so it can also be judged whether the video data belonging to the first image field _Fn is a film mode, and whether the first image field _Fn should be connected with the second image field Field Fn _-1 merges.

FIG. 2 , FIG. 3 , FIG. 4 and FIG. 5 are flow charts of the device in FIG. 1 performing film mode detection according to the technical spirit of the present invention. For the three programs in Figure 2, please refer to Figure 3, Figure 4 and Figure 5 respectively. The flow method is a method for judging whether a video data conforms to a specific video mode by using displacement estimation. The flow method receives three adjacent image fields from the video data as the front, middle, and rear image fields F _n-1 , F _n and F _n+1 respectively. As shown in Figure 2, after step 200, step 202 to step 216, step 302 to step 316 and step 402 to step 416 are respectively used to realize a displacement estimation and judgment step; after performing steps 216, 316 and 416, Then proceed to steps 218 and 220 to complete the flow of the method. A displacement estimation and judgment step includes the following steps:

Step 200: start the detection process of the present invention;

Step 202: Receive a field F _n and a field F _n-1 in the video data as its input;

Step 204: There are M ₁ pixels in the image field F _n , and the M ₁ pixels are numbered sequentially; and N ₁ is the number of a currently selected target pixel, indicating that the current target pixel is in the image field F _n N _1th pixel; set N ₁ =1;

Step 206: Perform displacement estimation for the target pixel, and estimate that the target pixel will fall at a certain position in the field Fn _-1 , so as to obtain a corresponding displacement estimation result, the displacement estimation result includes a corresponding A displacement vector MV_1 and a similarity S ₁ of ;

Step 208: Generate a first judgment value D 1 corresponding to the target pixel according to the magnitude relationship between the displacement vector MV_1 corresponding to each target pixel and the similarity S ₁ and a first critical value, the second judgment value D ₁ corresponding to the target pixel A judgment value _D1 represents whether the target pixel is displaced;

Step 210: Perform a single image field information step, according to the first judgment value D ₁ corresponding to each target pixel in the image field _Fn , to cumulatively generate the first count value C corresponding to the image field _{Fn 1} to represent the number of displaced target pixels in the field _Fn ;

Step 212: Determine whether the target pixel is the last pixel in the field F _n , that is, whether N ₁ is equal to M ₁ ; if yes, proceed to step 216; if not, proceed to step 214;

Step 214: Add 1 to the value of _N1 to become a new _N1 value, and the current target pixel is the _N1th pixel in the field _Fn ; proceed to step 206;

Step 216: According to the size relationship between the first count value C ₁ and a second critical value, a second judgment value D ₂ corresponding to the image field _Fn is generated, and the second judgment value D ₂ is a representative image Whether field F _n has image field displacement;

The above steps 206 to 208 can be used to realize a target pixel displacement estimation and judgment step, while steps 210 to 216 can realize a field displacement judgment step.

Steps 302 to 316 are similar to steps 202 to 216 , and the inputs received in step 302 are field F _n and field F _n+1 . Steps 402 to 416 are also similar to steps 202 to 216 , and the inputs received in step 402 are field F _n−1 and field F _n+1 . The next steps in this film mode detection process are as follows:

Step 218: Generate a judgment result according to the second judgment value D ₂ , the fourth judgment value D ₄ and the sixth judgment value D ₆ , the judgment result includes a seventh judgment value D ₇ and an eighth judgment value _D8 . The seventh judgment value _D7 represents whether the video data to which the field _Fn belongs is a video mode, for example, it is a 3-2 pull down video mode or a 2-2 pull down video mode. And the eighth judgment value _D8 represents whether the field _Fn should be merged with the field Fn _-1 or the field Fn ₊₁ ; and

Step 220: end.

The above is a preferred embodiment of the present invention, but in another implementation situation of the present invention, it is also possible to use only one displacement estimation and judgment step, that is, step 202 to step 216, in conjunction with step 200, step 218 and Step 220, and the input is the picture field F _n and the picture field F _n-1 so it can also be done to judge whether the video data to which the picture field F _n belongs is a film mode, and whether the picture field F _n should be the same as the picture field F _{n- 1} to merge.

Compared with the prior art, the present invention can more accurately determine the video mode to which a video data belongs, and can further select the correct de-interlacing method to correctly process the video data.

The above descriptions are only preferred embodiments of the present invention, and all equivalent changes and modifications made according to the claims of the present invention shall fall within the scope of the patent of the present invention.

Claims (10)

1. A film mode detection device for detecting a film mode of a video data, the video data is composed of a plurality of image fields, and the film mode detection device comprises: At least one displacement estimation and judgment module, the displacement estimation and judgment module includes: A pixel displacement estimation unit, receiving a first field and a second field in the video data, and based on a target pixel area of the first field and a plurality of pixel areas in the second field providing a displacement vector corresponding to the target pixel area; a field displacement judging unit, coupled to the pixel displacement estimation unit, to generate a second judging value corresponding to the first field according to the corresponding displacement vector in the first field; and A film mode detector is coupled to the at least one displacement estimation and judging module, receives the at least one second judging value, and generates a judging result. 2. The film mode detection device as claimed in claim 1, wherein the pixel displacement estimation unit comprises: A pixel displacement estimator obtains a displacement vector according to a target pixel area in the first field and a plurality of pixel areas in the second field. 3. The film mode detection device as claimed in claim 2, wherein the pixel displacement estimator outputs a similarity, the pixel displacement estimation unit comprises a first judger coupled to the pixel displacement estimator, when the The displacement vector represents when there is no displacement, and is used to generate the first judgment value based on comparing the similarity with a first critical value. 4. The film mode detection device as claimed in claim 1, wherein the image field displacement judging unit comprises: a field displacement estimator, used to accumulate the corresponding first judgment value to generate a count value corresponding to the first field; and A second judger, coupled to the field displacement estimator, is used to generate a second judgement value according to the count value. 5. A method for detecting a film mode of a video data, the video data is composed of a plurality of image fields, the method comprising: receiving a first field and a second field; comparing a target pixel area of the first image field with a plurality of pixel areas of the second image field, and obtaining a displacement vector and a similarity corresponding to the target pixel area; generating a first judgment value corresponding to the first image field according to the displacement vector and the similarity; and A second judgment value is generated according to the first judgment value. 6. The method of claim 5, wherein when the displacement vector represents no displacement, the first judgment value is generated by comparing the similarity with a first critical value. 7. The method as claimed in claim 5, wherein the step of generating the second judgment value comprises: accumulating the first judgment value to generate a count value corresponding to the first field; and A second judgment value corresponding to the first field is generated according to the count value and a second critical value. 8. A method for detecting a film mode of a video data, the video data is composed of a plurality of image fields, the method comprising: receiving a first field and a second field; comparing a target pixel area of the first field with multiple pixel areas of the second field, and obtaining a displacement vector corresponding to the target pixel area; generating a second judgment value corresponding to the first field according to the displacement vector; and A detection result is generated according to the second judgment value. 9. The method of claim 8, further comprising: accumulating the displacement vectors to generate a count value corresponding to the first field; and A second judgment value corresponding to the first field is generated according to the count value and a second critical value. 10. A film mode detection device for detecting a film mode of a video data, the video data is composed of a plurality of image fields, the device comprises: At least one displacement estimation and judgment module, the displacement estimation and judgment module includes: A pixel displacement estimation unit, receiving a first field and a second field in the video data, and based on a target pixel area of the first field and a plurality of pixel areas in the second field providing a displacement vector corresponding to the target pixel area and a similarity; and A field displacement judging unit, coupled to the pixel displacement estimation unit, generates a second judgment value corresponding to the first field according to the corresponding displacement vector and the similarity in the first field ;as well as A film mode detector is coupled to the at least one displacement estimation and judging module, receives the at least one second judging value, and generates a judging result.

Images