KR20040022013A - Vertical down scaling apparatus and method in the image processing system - Google Patents

Abstract

PURPOSE: A vertical down-scaling apparatus and method in a video signal processing system are provided to prevent the picture quality of an output image from being deteriorated due to aliasing. CONSTITUTION: A vertical down-scaling apparatus of a video signal processing system includes a pre-down-scaler(310) and a post-down-scaler(320). The pre-down-scaler down-scales an input video signal at a non-integer rate of a predetermined scale ratio. The post-down-scaler scales the down-scaled video signal at an integer rate of the predetermined scale ratio. The pre-down scaler filters the input video signal through linear interpolation using 2 lines. The pre-down-scaler includes line memories(311,313) for storing the input video signal, the first multiplier(312_1) for multiplying the video signal by the first coefficient, the second multiplier(312_2) for multiplying the video signal by the second coefficient, a coefficient generator for generating the first and second coefficients, an adder(314) for adding up the output signals of the first and second multipliers, an enable signal generator(316) for generating an enable signal, and a buffer(315) for buffering a signal output from the adder and outputting the buffered signal according to the enable signal.

Description

Vertical down scaling apparatus and method in the image processing system

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a vertical down scaling apparatus and method in a video signal processing system. In particular, the present invention relates to a video signal input system in a video signal processing system having a picture in picture (PIP) or picture outside picture (POP) function. Or a vertical downscaling apparatus and method used when processing for POP.

PIP and POP function outputs different video signal by dividing it into main image and sub image. Therefore, the video signal (screen for PIP or POP) output as the auxiliary image among the input video signals should be downscaled. Down scaling can be divided into horizontal down scaling and vertical down scaling.

The vertical down scaling apparatus used in the existing video signal processing system is shown in FIG. 1 illustrates a filtering method using n-1 line memories 101_1 to 101_n-1. That is, the input image data is sequentially stored in the first to n-th line memories 101_1 to 101_n. Since this vertical down scaling apparatus uses n lines, the coefficient generator 103 generates n filter coefficients (coefficients 1 to n). The n filter coefficients (coefficients 1 to n) are previously determined by a scale ratio.

The generated n filter coefficients are respectively transmitted to multipliers 102_1 to 102_n. The multipliers 102_1 to 102_n multiply coefficients provided from the coefficient generator 103 by image data inputted from the first to n-th line memories 101_1 to 101_n by the coefficients provided by the coefficient generator 103. do. Data output from the multipliers 102_1 to 102_n is added by the adder 104 and transmitted to the buffer 105. Data output from the adder 104 is data filtered according to the scale ratio.

The buffer 105 stores the filtered data transmitted from the adder 104 and transmits and stores the filtered data stored by the enable signal generated by the enable signal generator 106 to the memory 107. Let's do it. The enable signal generator 106 generates an enable signal by calculating a timing at which the filtered data is output according to a predetermined scale ratio.

However, in the above-described vertical down scaling apparatus, when the scale ratio is set to be larger than 1 / n, overlapping noises are generated in the output image and the image quality is deteriorated. When downscaling to 1 / (n + 1), n + 1 lines should be referenced. Since there are n reference lines, the reference information is insufficient, resulting in overlapping noise in the output image. Therefore, since a line memory must be provided in consideration of the maximum settable scale ratio, a lot of hardware is required.

2 is another example of a vertical down scaling apparatus used in a conventional video signal processing system. 2 is a method of accumulating and filtering image data input to one line memory. That is, when the down scale ratio is 1 / n, the n lines are stored in the line memory 204. At this time, the image data stored in the line memory 204 is multiplied by the coefficients 1 to n sequentially generated by the coefficient generator 201 to the image data sequentially input, and the filtering data previously stored in the adder 203 Store the added result. The operation of the line memory 204 and the coefficient generator 201 is controlled by the controller 208. The buffer 205 and enable signal generator 206 and memory 207 operate as shown in FIG. 1.

However, this downscaling device can be used only when the scale ratio is an integer multiple. If the scale ratio is non-integer, it is difficult to implement an algorithm to be set in the controller 208 to output the image without the overlapping noise.

SUMMARY OF THE INVENTION An object of the present invention is to provide a vertical down scaling apparatus capable of providing a PIP (or POP) screen without deterioration in image quality using a minimum line memory.

In order to achieve the above object, an apparatus according to the present invention includes: a vertical down scaling apparatus of a video signal processing system, the apparatus comprising: a free down scale unit which down scales an input video signal to a non-integer ratio of a set scale ratio; Preferably, the free-down scaler includes at least one post-down scaler for scaling down the downscaled video signal to an integer ratio of a set scale ratio.

Preferably, the pre-down scale unit is configured to filter an image signal input by a linear interpolation method using two lines.

The pre-down scale unit may include a line memory configured to store an input image signal in line units, a first multiplier determined according to a non-integer ratio, a first multiplier that multiplies an input image signal, and a second coefficient determined according to a non-integer ratio. A second multiplier for multiplying an image signal, a coefficient generator for generating a first coefficient and a second coefficient determined according to a non-integer ratio, an adder for adding signals output from the first multiplier and a second multiplier, and a non-integer ratio It is preferable to include an enable signal generator for generating an enable signal, and a buffer for buffering a signal output from the adder and outputting the signal according to the enable signal generated in the enable signal generator.

Preferably, the vertical down scaling apparatus has a configuration in which the number of filter taps is increased by the number of times provided when a plurality of post down scale units are provided.

The post-down scaler may include a coefficient generator that sequentially generates coefficients set according to an integer ratio, a multiplier that multiplies coefficients generated by the coefficient generator with data output from the free-down scale unit, data output from the multiplier, and previously filtered Adder to add data, update stored data with data output from adder, line memory to provide stored data to adder, buffer to buffer data output from adder, enable signal for buffer according to integer ratio It is preferable to include an enable signal generator.

The vertical down scaling apparatus may further include a multiplexer for selectively transmitting signals output from the plurality of post down scale units when a plurality of post down scale units are provided.

According to an aspect of the present invention, there is provided a method of vertical downscaling a video signal processing system, the method comprising: downscaling to a non-integer ratio of a scale ratio set for an input video signal; And scaling the video signal down-scaled to a non-integer ratio to an integer ratio of a set scale ratio.

In the scaling to the non-integer ratio, it is preferable to scale down by a linear interpolation method. The scaling by the integer ratio is preferably scaled down by a cumulative filtering method.

1 is an example of a conventional vertical down scaling apparatus used in a video signal processing system.

2 is another example of a conventional vertical down scaling apparatus used in a video signal processing system.

3 is a block diagram of a vertical down scaling apparatus according to a preferred embodiment of the present invention.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings.

3 is a block diagram of a vertical down scaling apparatus according to a preferred embodiment of the present invention. Referring to FIG. 3, the vertical down scaling apparatus includes a predown scale unit 310, a first postdown scale unit 320, a second post down scale unit 330, and a multiplexer 340. , And the memory 350.

When the desired down scale ratio is non-integer, the free down scale unit 310 down scales the video signal input at the non-integer ratio of the desired down scale ratio. For example, if you want to reduce 720 lines to 135 lines, the desired down scale ratio is '1 / 5.3'. Therefore, since the pre-down scale unit 310 down-scales the video signal input at the '1 / 0.3' ratio, the video signal of 720 lines is reduced to 675 lines.

The non-integer ratio is provided from a controller not shown. The control unit (not shown) separates the integer ratio and the non-integer ratio by analyzing the scale ratio requested by the user, and provides the non-integer ratio to the free down scale unit 310, and the integer ratio is the first post down scale unit 320 and The second post down scale unit 330 is provided.

As illustrated in FIG. 3, the pre-down scale unit 310 is configured to filter an image signal input by a linear interpolation method using two lines. That is, referring to FIG. 3, the free-down scale unit 310 may include a first line memory 311, a plurality of multipliers 312_1 and 312_2, a first coefficient generator 313, an adder 314, and a first buffer ( 315, a first enable signal generator 316.

The first line memory 311 stores an input video signal in line units. The multiplier 312_1 multiplies the coefficient 1 by the input video signal and outputs it. The multiplier 312_2 multiplies the coefficient 2 by the video signal output from the first line memory 311 and outputs the result. The first coefficient generator 313 provides coefficients 1 and 2 determined by the non-integer ratios applied to the multipliers 312_1 and 312_2, respectively. The method of determining the coefficients 1 and 2 according to the non-integer ratio applied by the first coefficient generator 313 may be implemented to use one of operating algorithms of the coefficient generator used in the existing vertical down scaling apparatus.

The signal output from the multipliers 312_1 and 312_2 is added at the adder 314. The signal added by the adder 314 is a signal filtered by two-line linear interpolation. The signal output from the adder 314 is transmitted to the first buffer 315. The first buffer 315 outputs a signal buffered by the enable signal generated by the first enable signal generator 316. The first enable signal generator 316 calculates a timing at which data is output from the first buffer 315 according to an applied non-integer ratio, and generates an enable signal based on the calculated result. The method of calculating timing according to the non-integer ratio may use one of the existing methods.

The first post down scale unit 320 down scales the input image signal according to the applied integer ratio. For example, in the case where it is desired to reduce an image signal of 720 lines to 135 lines, since the desired scale ratio is '1 / 5.33', the free down scale unit 310 downscales 720 lines to 675 lines as described above. When one image signal is provided, the first post down scale unit 320 downscales 675 lines to 1/5 to output an image signal having 135 lines. (Fig. 3 shows the case where the first and second post-down scale units 320 and 330 are connected and operated. Therefore, the coefficients used are also 1 to 2n. Isn't the coefficient 1-10?)

To this end, the first post down scale unit 320 may include a second coefficient generator 321, a multiplier 322, an adder 323, a second line memory 324, and a second buffer 325 as shown in FIG. 3. ), And a second enable signal generator 326.

The second coefficient generator 321 sequentially generates coefficients 1 to 2n determined according to the applied integer ratio. The method of determining the coefficients 1 to 2n according to the integer ratio uses one of the known methods. If the integer ratio is 5, the coefficients generated by the second coefficient generator 321 are 1-10. The generated coefficients 1 to 2n are transmitted to the multiplier 322.

The multiplier 322 multiplies the image generated from the second coefficient generator 321 by the video signal transmitted from the pre-down scale unit 310 to output the multiplied coefficient. The output multiplication result is transmitted to the adder 323. Adder 323 adds the signal sent from multiplier 322 and the previously added signal output from second line memory 324. The added result is transmitted to the second line memory 324 and the second buffer 325, respectively. Accordingly, the video signal transmitted from the free down scale unit 310 is cumulatively filtered. If the generated coefficients are 1 to 10, 10 lines of data are cumulatively filtered using the adder 323 and the second line memory 324.

The second buffer 325 buffers the signal transmitted from the adder 323 and outputs a signal buffered by the enable signal generated from the second enable signal generator 326. The second enable signal generator 326 calculates a timing at which data is output, as in the above-described first enable signal generator 316 according to the applied integer ratio, and generates an enable signal according to the calculated result. The signal output from the second buffer 325 is transmitted to the multiplexer (MUX) 340.

Meanwhile, as illustrated in FIG. 3, the second post-down scaling unit 330 includes a third coefficient generator 331, a multiplier 332, an adder 333, a third line memory 334, and a third buffer 335. ) And a third enable signal generator 336 to operate in the same manner as the first post down scale unit 320 described above. The cumulative filtered data output from the second post down scaler 330 is transmitted to another input terminal of the multiplexer 340.

The multiplexer 340 selectively transmits the scaled down image signal transmitted from the first post down scale unit 320 and the second post down scale unit 330 by a control signal transmitted from a controller (not shown). The control signal transmitted from the controller is generated based on the desired scale ratio.

The downscaled image signal output from the multiplexer 340 is stored in the memory 350. The memory 350 may be configured as an SDRAM.

In the case of FIG. 3 described above, two post-down scale units are used. However, only one post-down scale unit may be used or three or more may be implemented. If only one post-down scale unit is used, the coefficients generated by the coefficient generator included in the post-down scale unit are 1 to n coefficients. On the other hand, when three post down scale units are used, the coefficients generated by the coefficient generator included in the post down scale unit are 1 to 3 n coefficients. When the plurality of post-down scale units are implemented as described above, multiple filter lengths may be provided compared to a desired scale ratio, thereby preventing the output image from being degraded due to overlapping noise.

The vertical down scaling method according to the present invention can be implemented by a process based on the block diagram of the apparatus shown in FIG. That is, when the video signal is input, the process may include downscaling to a non-integer ratio of the set scale ratio and scaling a video signal downscaled to a non-integer ratio to an integer ratio of the set scale ratio. The scaling by the non-integer ratio may be implemented to scale down by a two-line linear interpolation method as in the pre-down scale unit 310 described with reference to FIG. 3. The scaling by the sound integer ratio may be implemented by downscaling by a cumulative filtering method such as the post-down scaler 320 of FIG. 3.

According to the present invention described above, when scaling an input video signal to a desired vertical down scale ratio, by dividing down the non-integer ratio and the integer ratio, the hardware is reduced by using a smaller number of line memories than the conventional vertical down scaling apparatus. It is advantageous in application to limiting systems. In addition, by providing a plurality of post-down scale parts, it is possible to extend the filter length to a desired down scale ratio, thereby preventing the output image from being degraded due to the overlapping noise (aliasing).

The present invention is not limited to the above-described embodiments, and variations of the present invention can be made by those skilled in the art within the spirit of the present invention. Therefore, the scope of claims in the present invention will not be defined within the scope of the detailed description will be defined by the claims below.

Claims (9)

In the vertical down scaling apparatus of a video signal processing system, A pre-down scaler for down-scaling the input video signal to a non-integer ratio of a set scale ratio; And at least one post-down scaler configured to scale the video signal down-scaled by the free-down scaler to an integer ratio of the set scale ratio. The vertical down scaling apparatus of claim 1, wherein the pre-down scale unit is configured to filter the input image signal by a linear interpolation method using two lines. The method of claim 1, wherein the free down scale unit, A line memory for storing the input video signal in line units; A first multiplier for multiplying the first coefficient determined according to the non-integer ratio and the input video signal; A second multiplier that multiplies the second coefficient determined according to the non-integer ratio and the input video signal; A coefficient generator for generating the first coefficient and the second coefficient determined according to the non-integer ratio, An adder for adding signals output from the first multiplier and the second multiplier, respectively; An enable signal generator for generating an enable signal in accordance with said non-integer ratio, and And a buffer for buffering the signal output from the adder and outputting the signal according to the enable signal generated by the enable signal generator. The vertical down scaling apparatus of claim 1, wherein the vertical down scaling apparatus has a configuration in which the number of filter taps is increased by a number of times provided when a plurality of post down scale units are provided. The method of claim 1, wherein the post down scale unit, A coefficient generator for sequentially generating coefficients set according to the integer ratio, A multiplier for multiplying coefficients generated by the coefficient generator with data output from the pre-down scale unit; An adder for adding data output from the multiplier and previously filtered data, A line memory for updating stored data with data output from the adder and providing the stored data to the adder; A buffer for buffering data output from the adder, And an enable signal generator for generating an enable signal for the buffer in accordance with the integer ratio. The vertical down scaling apparatus of claim 1, wherein the vertical down scaling apparatus further comprises a multiplexer for selectively transmitting signals output from the plurality of post down scale units when the plurality of post down scale units are provided. In the vertical down scaling method of a video signal processing system, Downscaling to a non-integer ratio of a scale ratio set for the input video signal; And scaling the video signal down-scaled by the non-integer ratio by an integer ratio of the set scale ratio. 8. The method of claim 7, wherein said scaling to non-integer ratios comprises downscaling by linear interpolation. 8. The method of claim 7, wherein said scaling to an integer ratio comprises downscaling with a cumulative filtering method.