KR100269389B1 - Apparatus for outline emphasis of digtal image - Google Patents

Abstract

PURPOSE: A method for emphasizing an outline of a digital image is provided to obtain a clear outline by extracting and emphasizing a boundary of a recovered image. CONSTITUTION: A buffer portion(300) receives a decoded image signal and delays the decoded signal during a predetermined time. An outline detection portion(400) obtains a difference signal between cells and obtains an outline of the image by comparing the difference signal with a reference value. A control portion(500) generates a control signal for selecting an emphasizing part of the outline. An outline signal generation portion(600) generates an outline emphasizing signal. A switching portion(700) outputs selectively an original image signal or the outline emphasizing signal.

Description

Contour highlighting device of digital image

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to image processing of a received digital broadcast signal, and more particularly, to an apparatus for emphasizing a digital image by extracting and emphasizing an image contour.

In digital broadcasting for multi-channel and high-definition broadcasting, compression technology is used to maintain image quality at a desired level and to reduce the amount of data. Redundancy) can be eliminated to achieve high compression ratios to reduce transmission time. The concept of spatial redundancy emerged from the fact that pixels adjacent to each other on a screen have almost similar values.

Since the compression process generally takes a lot of time when processing the original digital image by pixel, block the image by a certain unit, and then reduce the frequency from the spatial domain to reduce the correlation between neighboring blocks. A discrete cosine transform (DCT) process is performed to convert the frequency domain.

The block that has undergone the DCT conversion process has a low correlation between coefficient values obtained due to the bias corresponding to the spatial frequency and shows the phenomenon that energy is concentrated toward the low frequency region. That is, when looking at the frequency domain of the image by the DCT, it can be seen that there is a DC value in the upper left side as shown in FIG.

Therefore, the value of the high frequency region with relatively low energy is discarded for compression. This means throwing away high-frequency information, that is, parts where the change between pixels in the image is severe (a lot of edges: facial border, nose border, flying hair, etc.). Then, after encoding using the algorithm to compress with the value of the low frequency region is transmitted.

A general digital image decoder in a receiver for reconstructing a compressed image is shown in FIG. 1.

Referring to FIG. 1, a bit stream separation unit 101 that separates only video information from a bit stream that is multiplexed with video information, audio information, and additional information, and the video separated by the bit stream separation unit 101. A buffer 102 which temporarily stores information, a VLD unit which removes an overhead portion from the image information bit stream output from the buffer 102 and extracts only pure image information and performs variable-length decoding (VLD); 103) an inverse quantization unit 104 which inversely quantizes predetermined data obtained by the VLD unit 103 to obtain a DCT coefficient value, and an IDCT which obtains pixel values of an original image from the DCT coefficient values obtained by the inverse quantization unit 104. The unit 109 receives a video signal decoded by the VLD unit 103 and reads a screen of a specific position indicated by a motion vector value of a macro block from the frame memory 108. An adder 107 for outputting a decoded video signal by adding the outputs of the motion compensator 106, the IDCT unit 105, and the motion compensator 106 to produce a next screen, and an output of the adder 107. And a frame memory 108 that temporarily stores and outputs the same to the motion compensator 106.

In FIG. 1 configured as described above, the bit stream separation unit 101 reads a bit stream of compressed data output from a storage medium (CD, etc.) or a tuner unit of a receiver, and separates video information and audio information. The data is temporarily stored in the buffer 102 and then output to the VLD unit 103. The VLD unit 103 obtains only pure image information by separating various head information and overhead information from a bitstream of the image information output from the buffer 102, and the image information is variable-length coding. Variable-length decode since it is the correct information. That is, the VLD unit 103 obtains the quantized coefficient value, the quantizing parameter value and the motion vector value per macroblock by variable length decoding of pure image information, and then the obtained coefficient value and parameter are inversed. The quantizer 104 outputs the motion vector to the motion compensator 106.

The inverse quantization unit 104 performs an inverse quantization process of dividing the quantization coefficient value and the parameter output from the VLD unit 103 into a constant value, obtains the DCT coefficient value, and outputs the IDT unit 105 to the IDCT unit 105. The unit 105 restores the pixel value of the original image from the DCT coefficient value.

Also, the motion compensator 106 reads from the frame memory 108 a screen of a specific position of the first decoded intra picture (I picture) indicated by the motion vector value of a macroblock output from the VLD unit 103. Then, the image is generated and output to the adder 107. The adder 107 adds the pixel value of the original image output from the IDCT unit 105 and the next image generated by the motion compensator 106 to output a decoded image, that is, an image block which is an image before compression.

However, as described above, in order to remove spatial redundancy in the image compression process of the transmitting end, the original image is converted into a frequency distribution region as shown in FIG. 2 through a DCT transformation process, wherein the image block subjected to the DCT transformation process is a low frequency region. Energy is concentrated in the high frequency region, which is relatively low in energy, and is discarded for compression during quantization.

Therefore, when the compressed image is restored by the digital image decoder as shown in FIG. 1, since the high frequency component is already removed during the compression process, the boundary part of the image is not clear or aggregated, which may cause deterioration of image quality. .

SUMMARY OF THE INVENTION The present invention has been made to solve the above problems, and an object of the present invention is to provide an apparatus for emphasizing a digital image to obtain a clearer contour by emphasizing the value of the boundary part after extracting the boundary line of the reconstructed image. .

1 is a block diagram of a general digital video decoding apparatus

2 is a diagram illustrating a frequency domain of an image by DCT conversion.

3 is a block diagram illustrating a device for enhancing contour of a digital image according to the present invention;

4 is a detailed block diagram of FIG.

5 is a diagram illustrating an example of a contour discriminating position according to the present invention;

Explanation of symbols for main parts of the drawings

300: buffer sections 301 to 304,501: delay unit

400: contour detection unit 401, 403, 405: subtractor

402,404,406: Absolute value converting unit 407: Maximum value selecting unit

408 to 410: comparator 411: end gate

500: outline highlight selection unit 502: ora gate

600: video signal generator 700: switching unit

In order to achieve the above object, the digital image contour enhancement apparatus according to the present invention outputs an image signal before compression through a variable length decoding, inverse quantization, inverse DCT, and motion compensation when a compressed image signal is input. A buffer unit that receives the decoded image signal from the digital image decoding apparatus and sequentially delays the pixel for a predetermined pixel, obtains a difference signal between two adjacent pixels by receiving each pixel of the buffer unit, and compares the image signal with a preset reference value An outline detection unit for detecting an outline portion, an outline enhancement selection unit for outputting a control signal for selecting a portion to enhance the outline according to the outline detection result of the outline detection unit, an outline signal generator for generating an outline enhancement signal, and The original image signal or image which is switched according to the control signal of the contour enhancement selection unit and output from the buffer unit It characterized by configured by comprising a switch for selecting the contour enhancement signal output from the contour signal generator output.

The buffer unit may include first to fourth delayers configured to receive the decoded video signal and sequentially delay the pixels one by one.

The contour detection unit obtains a difference signal between the decoded image signal and the output of the first delay unit of the buffer unit, and then obtains an absolute value of the difference signal, an output of the first delay unit of the buffer unit, After obtaining the difference signal between the output of the second delayer and obtaining the difference signal between the output of the second delayer and the output of the third delayer, A third reference unit that takes an absolute value in the difference signal, a maximum value selection unit that receives the outputs of the first and third operation units, respectively, and selects and outputs a maximum value thereof; If the output value of the second calculation unit is smaller than the second reference value and smaller than the third reference value, it is determined that the pixel is the outline portion of the image between the pixel input to the second delay unit and the pixel output from the second delay unit. And it characterized by a comparison part configured to output a logic signal corresponding thereto.

The contour enhancement selector comprises a delay for delaying a clock by receiving the output of the contour detector and an OR gate for ORing the output of the contour detector and the output of the delay, the number of delays being an contour enhancement signal. It is characterized by the number of pixels to be emphasized.

The contour generator generates a signal for emphasizing the image of the contour portion detected by the contour detector as a preset reference value, the value of which is changed according to the degree of emphasis of the detected contour portion.

When a boundary line is extracted from the image data input by the contour emphasis apparatus, a signal for contour emphasis, not an input image, is output and the portion is emphasized, thereby improving the image quality of the digital image.

Other objects, features and advantages of the present invention will become apparent from the following detailed description of embodiments taken in conjunction with the accompanying drawings.

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

3 is a block diagram illustrating a contour enhancement apparatus for a digital image according to an embodiment of the present invention, in which a buffer unit 300 temporarily stores and outputs a decoded image block, and receives an output of the buffer unit 300 from a boundary of an image. Contour detection unit 400 for detecting an area (contour area), contour enhancement selection unit 500 for outputting a control signal for selecting a portion to be emphasized according to the detection result of the contour detection unit 400, contour of the contour The contour signal generator 600 for generating a shape, and the image signal contour-enhanced by the contour signal generator 600 or the output signal of the buffer unit 300 according to the control signal output from the contour enhancement selector 500. And a switching unit 700 for selectively outputting a signal.

As illustrated in FIG. 4, the buffer unit 300 receives a decoded image block and delays a pixel by a first delayer 301 and an output of the first delayer 301. Receives a second delayer 302 for delaying the delay, a third delayer 303 for delaying the output of the second delayer 302, and an output of the third delayer 303. And a fourth delay 304 for delaying one pixel.

In addition, the contour detection unit 400 of the first subtractor 401 and the first subtractor 401 obtaining a difference signal between the input image signal and the output of the first delayer 301 of the buffer unit 300. A first absolute value converter 402 which takes an absolute value at the output, a second subtractor 403 for obtaining a difference signal between the output of the first delayer 301 and the output of the second delayer 302, and A third absolute value converting unit 404 taking an absolute value at the output of the second subtractor 403, a third obtaining a difference signal between the output of the second delayer 302 and the output of the third delayer 303; The output of the subtractor 405, the third absolute value converter 406, and the first and third absolute value converters 402 and 406, which take an absolute value at the output of the third subtractor 405, are duplicated. A maximum value selector 407 for selecting a value, a first comparator 408 for comparing the output value of the maximum value selector 407 with a first reference value ref1 and outputting a result thereof, and the second absolute value Second and third comparators 409 and 410 for comparing the output value of the affected part 404 with the second and third reference values ref2 and ref3 and outputting the results, respectively, and the first to third comparators 408 to 410. It is composed of an AND gate 411 which outputs '1' only when all of the outputs thereof become logic '1'. Here, the first to third reference values Ref1 to Ref3 are preset values and are determined according to the boundary range of the image to be determined as the outline. In other words, it is different depending on whether only a portion having a strong boundary is used as the contour emphasis portion or a portion having a non-boundary boundary as the contour emphasis portion.

In addition, the contour emphasis selector 500 receives the output of the contour detector 400 to delay the clock by one clock, the fifth delayer 501, and the output of the contour detector 400 and the fifth delayer ( When the output of 501 is received and any one has a logic '1', the OR gate 502 outputs a logic '1'. Here, the number of the retarders is determined by how many pixels the contour emphasis is horizontally. In other words, the width of the contour is determined. For example, when one retarder 502 is used as shown in FIG. 4, the contour emphasis is also performed on the adjacent pixels of the pixels detected as the contour portions.

In addition, the contour signal generator 600 is configured as a buffer for receiving and outputting the contour emphasis signal ref4 for the contour. The contour enhancement signal Ref4 is a signal in which a luminance signal or a color signal or a luminance signal and a color signal are synthesized. The contour enhancement signal Ref4 is a signal generated to arbitrarily create an image of the detected contour portion regardless of the original image signal. The value varies depending on the emphasis of the part.

In the present invention configured as described above, the image data, which is decoded and output as shown in FIG. 1, is input to the buffer unit 300 and temporarily stored in a delayed form by one pixel in each of the delayers 301 to 304. It is also output to 400.

The contour detector 400 uses the pixel data temporarily stored in the buffer unit 300 to output a pixel A1 and a second delayer which are delayed by one pixel from the boundary portion of the image, that is, the first delayer 301. At 302, it is detected whether the boundary between the pixels A2 output by being delayed by one pixel again. To this end, the subtractor 401 of the contour detector 400 obtains a difference signal A0-A1 between the input signal A0 of the buffer unit 300 and the output signal A1 of the first delay unit 301, 1 The absolute value converter 402 takes an absolute value for the difference signal A0-A1 (| A0-A1 |). In the second subtractor 403, the difference signal A1-A2 between the output signal A1 of the first delayer 301 of the buffer unit 300 and the output signal A2 of the second delayer 302. The absolute value of the difference signal A1-A2 is obtained by the second absolute value converting unit 404 (| A1-A2 |), and the third subtractor 405 outputs the second delay unit 302. After obtaining the difference signal A2-A3 between the signal A2 and the output signal A3 of the third delayer 303, the third absolute value converter 406 determines the absolute value of the difference signal A2-A3. Is taken (| A2-A3 |).

The maximum value selection unit 407 of the contour detection unit 400 outputs the output of the first absolute value conversion unit 402 (| A0-A1 |) and the output of the third absolute value conversion unit 406 (| A2- Compares A3 |) and selects the largest value of the two absolute values, that is, the maximum value, and outputs it.

The first comparator 408 of the contour detector 400 compares the output of the maximum value selector 407 with the first reference value Ref1 so that the output of the maximum value selector 407 becomes the first reference value Ref1. If greater, output a logic '0' signal, and if smaller, output a logic '1' signal. In addition, the second comparator 409 compares the output of the second absolute value converter 404 with the second reference value Ref2 to determine the output (| A1-A2 |) of the second absolute value converter 404. If greater than the second reference value Ref2, a signal of logic '1' is output, and if smaller, a signal of logic '0' is output. The third comparator 410 compares the output of the second absolute value converter 404 with the third reference value Ref3, so that the output of the second absolute value converter 404 (| A1-A2 |) If greater than the third reference value Ref3, a signal of logic '0' is output, and if smaller, a signal of logic '1' is output.

In this case, the outputs of the first to third comparators 408 to 410 are input to the AND gate 411, and the AND gate 411 performs an AND on the input signals to output the contour boundary determination signal of the image signal. The AND gate 411 outputs a signal of logic '1' only if the input signals are all signals of logic '1'. That is, as shown in FIG. 5, the output of the maximum value detector 407 is smaller than the first reference value Ref1 (| A0-A1 | <Ref1 and | A2-A3 | <Ref1), and the second absolute value converter If the output (| A1-A2 |) of 404 is larger than the second reference value Ref2 and smaller than the third reference value Ref3 (Ref2 < A1-A2 | < Ref3), then the output of the AND gate 411 is It becomes a signal of logic '1' and determines that the delayed signal A1 and signal A2 are the outline portions of the image.

The contour detection signal output from the AND gate 411 of the contour detector 400 is input to the fifth delayer 501 of the contour enhancement selector 500 to output a clock delayed signal, and the fifth delayer. The input signal and the output signal of 501 are respectively input to the OR gate 502. The OR gate 502 outputs a logic '1' signal when either the output signal of the AND gate 411 or the signal delayed by one clock by the fifth delayer 501 is a logic '1' signal. This signal is input to the switching unit 700 as a signal for selecting a portion to emphasize the outline.

On the other hand, the contour signal generator 600 buffers the preset reference signal Ref4 and outputs the signal to emphasize the contour.

The switching unit 700 is switched by a control signal for selection output from the contour enhancement selecting unit 500 and is made of a signal or buffer unit 300 for contour enhancement output from the contour generation unit 600. 4 A signal output from the delay unit 304, that is, an original video signal is selectively output.

As described above, according to the contour enhancement apparatus of the digital image according to the present invention, by determining the contour portion of the decoded image signal and outputting a predetermined contour enhancement signal to a portion where a sudden change of the image occurs, the DCT at the transmitting end. By compensating for a phenomenon in which the boundary portion of the reconstructed image is not cleared by the process, the digital image screen which is compressed and then reconstructed is more sharp.

Claims (6)

In the digital video decoding apparatus, if a compressed video signal is input, it is output as a video signal before compression through variable length decoding, inverse quantization, inverse DCT, and motion compensation. A buffer unit which receives the decoded video signal and sequentially delays the pixel for a predetermined pixel; An outline detection unit which receives each pixel of the buffer unit, obtains a difference signal between two adjacent pixels, and detects an outline portion of an image by comparing with a preset reference value; A controller for outputting a control signal for selecting a portion to be emphasized according to a contour detection result of the contour detector; An outline signal generator for generating an outline emphasis signal; And a switching unit which selects and outputs an original image signal outputted from the buffer unit or an outline enhancement signal outputted from the contour signal generator, according to a control signal of the controller. The method of claim 1, wherein the buffer unit And a first to fourth delayers which receive the decoded video signal and sequentially delay the pixels one by one. The method of claim 1, wherein the contour detection unit A first calculating unit which obtains a difference signal between the decoded and output image signal and the output of the first delay unit of the buffer unit and takes an absolute value in the difference signal; A second operation unit which obtains a difference signal between the output of the first delay unit and the output of the second delay unit and takes an absolute value to the difference signal; A third calculating unit which obtains a difference signal between the output of the second delay unit and the output of the third delay unit of the buffer unit and takes an absolute value in the difference signal; A maximum value selecting unit which receives the outputs of the first and third calculating units, respectively, and selects and outputs a maximum value thereof; And a comparator configured to detect an outline part of the image by comparing the output of the second calculator and the output of the maximum value selector with preset reference values, respectively. The method of claim 3, wherein the comparison unit When the output of the maximum selector is smaller than the first reference value which is set in advance, and the output of the second calculator is larger than the second reference value and smaller than the third reference value, the pixel input to the second delay unit of the buffer unit and the second delay unit are output. It is determined that the pixel is a contour portion of the simplified image and outputs a logic signal corresponding thereto. The method of claim 1, wherein the control unit And a delay gate for delaying a clock by receiving the output of the contour detection unit, and an OR gate for ORing the output of the contour detection unit and the output of the delay unit, and the number of the delay units is the number of pixels to be emphasized by the contour enhancement signal. Contour enhancement device of a digital image, characterized in that according to. The method of claim 1, wherein the contour generating unit And generating a signal for emphasizing the image of the contour portion detected by the contour detection unit as a preset reference value, the value of which is changed according to the degree of emphasis of the detected contour portion.