JP4433949B2 - Image processing apparatus and method - Google Patents

Description

  The present invention relates to an image processing apparatus for progressively converting a signal in which a conversion signal converted into an interlace by 3-2 pulldown, 2-2 pulldown, or the like and a normal interlace signal of, for example, 60 fields / second are mixed. It relates to that method.

  Standard television signals such as NTSC signals and high vision signals are interlaced signals. 7A and 7B are diagrams showing a scanning line structure. FIG. 7A shows an interlace signal, FIG. 7B shows a progressive signal, and FIG. 7C shows an interlace signal converted into a progressive signal by scanning line interpolation. Signal. In FIG. 7, a circle indicates a scanning line, and a cross indicates an interpolated scanning line.

  In FIG. 7, the vertical direction V is the vertical direction of the screen, and the horizontal direction t is the time direction. As shown in FIG. 7A, the interlace signal is composed of two fields in which one frame is shifted in the time and vertical directions. On the other hand, the progressive signal has no deviation in the scanning line structure as shown in FIG. In an interlace signal, when a high frequency component in the vertical direction of an image increases, interlace interference such as line flicker occurs. On the other hand, there is no interlace interference in the progressive signal.

  Therefore, as shown in FIG. 7C, there is a processing method for removing interlace interference by interpolating a scanning line of a portion thinned out by interlacing with a surrounding scanning line and converting it into a progressive signal. is there. Such a processing method is called progressive conversion or double-density conversion.

  Conventionally, scanning line interpolation for progressive conversion is performed by motion adaptive interpolation processing. That is, as shown in FIG. 8, when the image is still, inter-field interpolation is performed with the average value of the signals PA and PB representing the pixels in the preceding and following fields as the signal PQ representing the new pixel indicated by the x mark. To generate a new scan line. On the other hand, when the image is moving, a new scanning line is generated by performing intra-field interpolation using an average value of the signals PC and PD representing the upper and lower pixels as a signal PQ representing a new pixel indicated by a cross. For this reason, when the image is stationary, a good conversion image quality with little aliasing distortion and high resolution can be obtained. However, when the image is moving, it becomes a deteriorated conversion image quality with much aliasing distortion and low resolution.

  Here, when the input signal to be converted to a progressive signal is a signal converted to interlace by 3-2 pulldown or 2-2 pulldown, an image is obtained by adopting a method different from the motion adaptive interpolation processing. Even if it moves, good conversion image quality can be obtained. Here, 3-2 pull-down refers to frame rate conversion as shown in FIG. Specifically, progressive signals A, B, C... Such as film data of 24 frames / second are used as interlace signals a, a ′, a, b ′, b, c of NTSC system of 60 fields / second. Used as a method for converting to ', c, c'. In the figure, the presence or absence of “′” indicates the difference between the odd field and the even field. On the other hand, the 2-2 pulldown refers to frame rate conversion as shown in FIG. More specifically, progressive signals A, B, C... Such as film data of 30 frames / second are converted into interlace signals a, a ′, b, b ′, c, c of NTSC system of 60 fields / second. Used as a method for converting to '...'.

  As shown in FIGS. 9 and 10, in the 3-2 pulldown, an image that was originally 1 frame is distributed into 3 or 2 fields, and in the 2-2 pulldown, an image that was originally 1 frame is 2 fields. It is distributed to. Therefore, if the 3-2 pattern or 2-2 pattern of the input signal converted to interlace by 3-2 pull-down or 2-2 pull-down is known, the image is stopped in the adjacent field generated from the same one-frame image. Regardless of movement, field interpolation can be performed and converted to a progressive signal. Here, in the field interpolation, although the interpolation method is different, a new scanning line is obtained by setting the previous field signal PA or the subsequent field signal PB to a signal PQ representing a new pixel, similarly to the interfield interpolation shown in FIG. Therefore, a good conversion image quality with little aliasing distortion and high resolution can be obtained.

  By the way, a normal 60 field / second interlace signal may be inserted by editing into a signal converted into an interlace by 3-2 pulldown or 2-2 pulldown. When such an interlace signal is converted into a progressive signal, a field generated from the same one-frame image does not exist in an adjacent field for a normal 60 field / second image portion, particularly a moving image. There is a problem in that an optimal progressive signal cannot be obtained by field interpolation and the image quality deteriorates. For example, as shown in FIG. 11, when an input signal in which an image portion indicated by a round object of 60 fields / second is inserted into a signal converted to interlace by 2-2 pulldown is converted into a progressive signal, Since a round object is moving as an interlace signal of field / second, when a progressive signal is obtained by field interpolation applied to 2-2 pull-down, the image portion is doubled, resulting in a large image quality. You will lose.

  In order to solve such problems and the like, the following Patent Document 1 describes an image in which an image converted to interlace by 3-2 pulldown or 2-2 pulldown and a normal 60 field / second image are mixed. Even if it exists, the technique of converting into favorable progressive without image quality degradation is disclosed. In the technique described in Patent Document 1, a post-field interpolation signal using a scanning line signal at the same position as an interpolated scanning line in a field located temporally behind the current field for scanning line interpolation. And a previous field interpolation signal using a scanning line signal at the same position as the interpolated scanning line of the field positioned temporally before the current field for scanning line interpolation, and a rear field interpolation signal. The signal indicating the minimum value is selected from the inter-frame matching signal that is the absolute value of the difference signal of the previous field interpolation signal, and is set as the motion signal K. Then, in accordance with the magnitude of the motion signal K, a field interpolation signal obtained by mixing the subsequent field interpolation signal and the previous field interpolation signal, and scanning lines above and below the interpolated scanning line in the current field are obtained. The mixing ratio with the interpolation signal in the current field generated by the addition is changed to obtain an optimum signal.

JP 2000-78535 A

  However, in the technique described in Patent Document 1 described above, since determination is made by comparing data of pixels having different fields, that is, having different times, the value of the same pixel is used when the inserted image portion is moving. Is different, and the process prioritizes intra-field interpolation. For this reason, there has been a problem that the image quality of the portion other than the inserted image portion that can be converted into progressive progressively without deterioration in image quality deteriorates.

  The present invention has been proposed in view of such a conventional situation. For example, a normal interlace of 60 fields / second is converted into a converted signal converted into an interlace by 3-2 pulldown, 2-2 pulldown, or the like. Even when the signal is inserted, that is, the interlace signal converted so that the frame rate is matched by rearranging the screen of a continuous original image such as a film based on a predetermined sequence, To provide an image processing apparatus and method capable of satisfactorily converting progressively without deterioration in image quality in both the inserted portion and other portions even when a normal interlace signal is inserted With the goal.

In order to achieve the above-described object, the image processing apparatus according to the present invention arranges the screen of the original image based on a predetermined sequence, and performs 3-2 pull-down so as to match the frame rate of the input video signal. Or an image processing apparatus that progressively converts an interlace signal that includes a conversion signal converted into an interlace by 2-2 pulldown as the video signal, and is located behind the current field in time. Progressive field interpolation by field interpolation from a scan line at the same position as the interpolated scan line, or from the scan line located before the current field in time and at the same position as the interpolated scan line A field interpolation signal generating means for generating a signal and a target pixel in the field interpolation signal A double image judging means for judging whether or not the pixel constitutes a double image portion of the signal, and intra-field interpolation from scanning lines positioned above and below the interpolated scanning line in the current field, Alternatively, a scan line that is located behind the current field and located at the same position as the interpolated scan line, and a scan line located before the current field and located at the same time as the interpolated scan line. Interpolation signal generation means for generating a progressive interpolation signal from the line by inter-field interpolation, and the double image determination means includes, for the target pixel, a straight line including the target pixel and orthogonal to the scanning line. If the correlation between pixels on one scanning line is larger than the correlation between pixels on adjacent scanning lines, it is determined that the pixel constitutes the double image portion, The field interpolated signal corresponding to the pixel of interest generated by the field interpolated signal generating means is replaced with the interpolation signal generated by the intra-field interpolation or inter-field interpolation corresponding to the pixel of interest.

  Further, when the double image determination unit determines that the target pixel is a pixel constituting the double image portion, the field interpolation signal corresponding to the target pixel generated by the field interpolation signal generation unit Is replaced with a predetermined replacement signal corresponding to the target pixel.

  Here, the image processing apparatus is located at the same position as the interpolated scan line that is located behind the current field by intra-field interpolation from the scan line located above or below the interpolated scan line in the current field Interpolating signal generating means for generating a progressive interpolation signal by inter-field interpolation from the scanning line and the scanning line located before the current field in time and located at the same position as the interpolated scanning line. In this case, when the double image determination unit determines that the target pixel is a pixel constituting the double image portion, the double pixel determination unit interpolates the target pixel in the field interpolation signal as a predetermined replacement signal. Replace with a pixel at the same position as the target pixel in the signal.

In order to achieve the above-described object, the image processing method according to the present invention arranges the screen of the original image based on a predetermined sequence so as to match the frame rate of the input video signal. An image processing method for progressively converting an interlaced signal including a conversion signal converted into an interlace by 2-pulldown or 2-2 pulldown as the video signal, which is positioned behind the current field in time. In the progressive field, the scanning line at the same position as the interpolated scanning line of the current field, or the scanning line located before the current field in time and by the field interpolation from the scanning line at the same position as the interpolated scanning line. A field interpolation signal generation step for generating an interpolation signal, and a target pixel in the field interpolation signal is the field A double image determination step for determining whether or not the pixel constitutes a double image portion of the interpolation signal, and intra-field interpolation from scanning lines positioned above and below the interpolated scanning line in the current field; Or a scan line that is located behind the current field in time and located at the same position as the interpolated scan line, and that is located before the current field in time and located at the same position as the interpolated scan line. An interpolation signal generation step of generating a progressive interpolation signal from the scanning line by inter-field interpolation. In the double image determination step, for the target pixel, a straight line that includes the target pixel and is orthogonal to the scanning line Among the upper pixels, if the correlation between the pixels on the scanning line one scan line apart is larger than the correlation between the pixels on the adjacent scanning line, it is determined that the pixel constitutes the double image portion. And, replacing the fields in the field interpolated signal corresponding to the pixel of interest generated by the inter-signal generating step interpolating, by the interpolation signal generated by the intra-field interpolation or inter-field interpolation corresponding to the pixel of interest .

  Further, in the double image determining step, when it is determined that the target pixel is a pixel constituting the double image portion, the field interpolation corresponding to the target pixel generated in the field interpolation signal generation step is performed. The signal is replaced with a predetermined replacement signal corresponding to the target pixel.

  Here, this image processing method is the same position as the interpolated scanning line that is located behind the current field by interpolating from the scanning line positioned above and below the interpolated scanning line in the current field or temporally behind the current field. And an interpolation signal generation step of generating a progressive interpolation signal by inter-field interpolation from the current scan line and a scan line located before the current field in the same position as the interpolated scan line. In this case, in the double image determining step, when it is determined that the target pixel is a pixel constituting the double image portion, the target pixel in the field interpolation signal is used as a predetermined replacement signal. Replace with a pixel at the same position as the target pixel in the interpolation signal.

  In such an image processing apparatus and its method, a signal in which a converted signal converted to interlace by 3-2 pulldown or 2-2 pulldown and a normal interlace signal of, for example, 60 fields / second are mixed is included in the field. When converting to a progressive field interpolation signal by interpolation, a double image portion in the field interpolation signal is detected on a pixel-by-pixel basis, and motion adaptive interpolation is performed for pixels constituting the double image portion. It replaces with the same pixel in the other signal obtained by etc.

  According to the image processing apparatus and method of the present invention, a signal in which a converted signal converted to interlace by 3-2 pulldown or 2-2 pulldown and a normal interlace signal of, for example, 60 fields / second are mixed. Is converted into a progressive field interpolation signal by field interpolation, the converted signal and a normal interlace signal can be reliably discriminated in pixel units. Further, whether or not the pixel of interest in the field interpolation signal constitutes a double image portion is determined on a pixel-by-pixel basis. Based on the determination result, motion adaptive interpolation is performed on the pixels constituting the double image portion. By replacing with the same pixel in other signals obtained by the above, etc., for example, both the part where the normal interlace signal of 60 fields / second is inserted and the other part are converted to progressive well without deterioration in image quality It becomes possible to do.

  Hereinafter, specific embodiments to which the present invention is applied will be described in detail with reference to the drawings. In this embodiment, even when a normal interlace signal of 60 fields / second is inserted in the converted signal converted into the interlace by the 3-2 pulldown or the 2-2 pulldown, The present invention is applied to an image processing apparatus and method that can convert progressively satisfactorily without deterioration in image quality for both the insertion portion and other portions.

  First, an example of a schematic configuration of the image processing apparatus according to the present embodiment is shown in FIG. As shown in FIG. 1, the image processing apparatus 100 according to the present embodiment includes a front image processing unit 10, a progressive conversion unit 11, a display device driving circuit 12, and a display device 13.

  The front image processing unit 10 inputs image signals from various signal sources such as an NTSC signal, a PAL signal, and an HDTV signal from a BS digital tuner. The format of the image signal is an interlace signal such as 525i (interlace signal with 525 lines), 625i, 1125i, or the like.

  The progressive converter 11 converts the 525i signal into a 525p signal (progressive signal with 525 lines), the 625i signal into a 625p signal, and the 1125i signal into a 1125p signal. At this time, the progressive conversion unit 11 will be described later even when a normal signal of 60 fields / second is inserted into the signal converted to interlace by 3-2 pulldown or 2-2 pulldown. Both the insertion part and the other part can be converted into progressive well without deterioration in image quality. The progressive conversion unit 11 supplies the obtained progressive signal to the display device drive circuit 12.

  The display device driving circuit 12 drives the display device 13 to display the progressive signal supplied from the progressive conversion unit 11. As this display device 13, various display devices such as a cathode ray tube, a liquid crystal display, and a plasma display can be used.

  Note that the display device drive circuit 12 may include a resolution conversion circuit that converts a standard resolution or low resolution image into a high resolution image including a high frequency component that is not included therein. Such resolution conversion circuits are described in, for example, Japanese Patent Application Laid-Open Nos. 7-193789 and 11-55630.

  Next, FIG. 2 shows an example of the internal configuration of the progressive conversion unit 11 described above. The interlace signal input from the front image processing unit 10 is supplied to a plurality of blocks of the progressive conversion unit 11 as a current signal. This current signal is delayed by a time corresponding to one field by the field delay unit 20 and converted into the past one signal, and further delayed by a time corresponding to one field by the field delay unit 21 and converted into the past two signals.

  The motion detection unit 22 performs motion detection using the current signal, the past 1 signal, and the past 2 signal. For determination of motion detection, a motion detection determination history stored in the memory 23 as to whether the pixel has been moving or stationary in the past is used. The motion detection unit 22 supplies the motion detection result to the motion adaptive interpolation signal generation unit 24. The motion adaptive interpolation signal generation unit 24 performs inter-field interpolation or intra-field interpolation based on the motion detection result, and generates a progressive motion adaptive interpolation signal. Specifically, when the image is stationary, a new scanning line is generated by performing inter-field interpolation using the average value of the pixels in the preceding and following fields as a new pixel. On the other hand, when the image is moving, a new scanning line is generated by performing intra-field interpolation using the average value of the upper and lower pixels as a new pixel. The motion adaptive interpolation signal generation unit 24 supplies the generated motion adaptive interpolation signal to the interpolation signal selection unit 26 and the pull-down error detection unit 28.

  The pull-down detection unit 25 detects 3-2 pull-down or 2-2 pull-down from the current signal, the past 1 signal, and the past 2 signal. Specifically, when the input signal is a signal converted to interlace by 3-2 pulldown or 2-2 pulldown, a field generated from an image of the same frame always exists in an adjacent field. By detecting the presence or absence of motion (correlation between fields), 3-2 pulldown or 2-2 pulldown can be detected. Note that, even if a normal signal is included in the signal converted to interlace by this 3-2 pulldown or 2-2 pulldown, the presence / absence of correlation between fields is determined by whether or not it is equal to or greater than a predetermined threshold value. Thus, detection is possible. The pull-down detection unit 25 supplies the pull-down detection result to the interpolation signal selection unit 26 and the pull-down error detection unit 28.

  The interpolation signal selection unit 26 determines an interpolation signal based on the motion adaptive interpolation signal, the current signal and the past two signals, and the pull-down detection result. That is, when the input signal is a signal converted to interlace by 3-2 pulldown or 2-2 pulldown, the field generated from the image of the same frame always exists in the adjacent field. Then, a field signal (current signal or past two signals) generated from the image of the same frame is supplied to the double speed conversion unit 27 as an interpolation signal. In addition, when the input signal is not a signal converted to interlace by 3-2 pulldown or 2-2 pulldown, the interpolation signal selection unit 26 uses the output signal of the motion adaptive interpolation signal generation unit 24 as an interpolation signal as a double speed conversion unit. 27. The double speed conversion unit 27 converts the interpolated signal and the past one signal into a progressive field interpolation signal by alternately reading at a speed twice that of the input signal, and the field interpolation signal is converted into a pull-down error detection unit. 28.

  Here, the conventional progressive conversion unit supplies the field interpolation signal to the display device driving circuit as it is. However, the signal converted to interlace by 3-2 pulldown or 2-2 pulldown is converted to 60 fields / When the signal into which the second signal is inserted is an input image, the inserted image portion is doubled, and there is a possibility that the image quality is greatly impaired.

  Therefore, the progressive conversion unit 11 in this embodiment includes a pull-down error detection unit 28 subsequent to the double-speed conversion unit 27. If the field interpolation signal is evaluated and a double image error is detected, the field conversion is performed. The interpolation signal is exchanged for a motion adaptive interpolation signal. In particular, the pull-down error detection unit 28 detects an error from information on surrounding pixels in units of pixels, and when a double image error is detected, replaces the field interpolation signal with a motion adaptive interpolation signal in units of pixels. . The detection of this double image is determined by the pull-down detection unit 25 that the input signal is a signal converted to interlace by 3-2 pull-down or 2-2 pull-down, and the interpolation signal selection unit 26 determines the current signal. Alternatively, it is performed only when the past two signals are used as interpolated pixels.

  An example of the internal configuration of the pull-down error detection unit 28 is shown in FIG. The field interpolation signal supplied from the double speed converter 27 is delayed by one line by the line delay unit 30. Further, the signals are delayed by one line by the line delay units 31, 32, and 33, respectively, and finally A, B, C, D, and E signals are obtained. The positional relationship between the signals A, B, C, D, and E is as shown in FIG. A dotted line in the figure is a scanning line generated by field interpolation, and a straight line in the figure is a scanning line as it is as an input signal.

  Here, in the case of a signal converted to interlace by 3-2 pull-down or 2-2 pull-down, the adjacent lines after progressive conversion were originally the same frame, and therefore should have a strong correlation. Conversely, in the case of a signal of 60 fields / second, the correlation should be stronger in the upper and lower lines at the time of interlacing, that is, the interval between two lines at the time of progressive, than the line correlation between the upper and lower sides. The pull-down error detection unit 28 having the configuration shown in FIG. 3 uses this difference in correlation to discriminate between a signal converted to interlace by 3-2 pull-down or 2-2 pull-down and a signal of 60 fields / second. .

  The difference absolute value calculation units 34 to 37 calculate the difference absolute value of the pixel data between adjacent lines. That is, the difference absolute value calculation unit 34 calculates the difference absolute value between the signal A and the signal B, and the difference absolute value calculation unit 35 calculates the difference absolute value between the signal B and the signal C. Similarly, the difference absolute value calculation unit 36 calculates the difference absolute value between the signal C and the signal D, and the difference absolute value calculation unit 37 calculates the difference absolute value between the signal D and the signal E. The one-line difference absolute value average calculation unit 38 obtains an average value of the difference absolute values calculated by the difference absolute value calculation units 34 to 37 and supplies the average value to the flip-flop (FF) 39. When the average value of the absolute value of one line difference is supplied for the surrounding five pixels by the flip-flops 39 to 42, the surrounding average calculation unit 43 calculates the average value for the surrounding five pixels and doubles this average value. The image is supplied to the image determination unit 53.

  On the other hand, the difference absolute value calculation units 44 to 46 calculate the difference absolute value of the pixel data between the lines separated by one line. That is, the difference absolute value calculation unit 44 calculates the difference absolute value between the signal A and the signal C, and the difference absolute value calculation unit 45 calculates the difference absolute value between the signal B and the signal D, thereby calculating the difference absolute value calculation. The unit 46 calculates the absolute difference between the signal C and the signal E. The two-line difference absolute value average calculation unit 47 obtains an average value of the difference absolute values calculated by the difference absolute value calculation units 44 to 46 and supplies the average value to the flip-flop 48. When the average value of the two-line difference absolute value is supplied for the surrounding five pixels by the flip-flops 48 to 51, the surrounding average calculation unit 52 calculates the average value for the surrounding five pixels and doubles the average value. The image is supplied to the image determination unit 53.

  The double image determination unit 53 compares the average value supplied from the peripheral average calculation unit 43 with the average value supplied from the peripheral average calculation unit 52, and if the latter is smaller, that is, between lines separated by one line. If the correlation is stronger, it is determined that a double image is generated. The double image determination unit 53 supplies the determination result to the output selection unit 55.

  The output selection unit 55 is supplied with the motion adaptive interpolation signal and the C signal delayed by two lines by the two-line delay unit 54 and the pull-down detection result. When the output selection unit 55 receives from the double image determination unit 53 that the double image has been generated, the output selection unit 55 uses the pixel of the C signal (field interpolation signal) as the pixel at the same position in the motion adaptive interpolation signal. To avoid the occurrence of a double image error.

  Another example of the internal configuration of the pull-down error detection unit is shown in FIG. As in FIG. 3, the field interpolation signal supplied from the double speed conversion unit 27 is delayed by one line by the line delay unit 60. Further, the signals are delayed by one line by the line delay units 61, 62 and 63, respectively, and finally the signals A, B, C, D and E are obtained.

  The average value calculation unit 64 calculates the average value of the signals A, B, C, D, and E, and the binarization unit 65 converts the signals A, B, C, D, and E into the binary signals A ′ and B ′. , C ′, D ′, and E ′. Specifically, if the pixel data is larger than the average value of the signals A, B, C, D, and E, for example, “1” is set, and if the pixel data is smaller, for example, “0” is set. The double image determination unit 66 determines that a double image is generated when the binary pattern for five lines matches a predetermined pattern.

  An example of a binary pattern for five lines is shown in FIG. As shown in FIG. 6, when the binary pattern for five lines matches the pattern (1, 0, 1, 0, 1) or (0, 1, 0, 1, 0), that is, one line apart. When the correlation between the lines is stronger, it is determined that there is a double image, and the determination result is supplied to the flip-flop 67. When the determination result of the double image is supplied by the flip-flops 67 to 70 for the surrounding five pixels on the same scanning line, the peripheral comprehensive double image determination unit 71 determines the double image from the determination result for the five surrounding pixels. Is comprehensively determined, and the determination result is supplied to the output selection unit 73. For example, the peripheral comprehensive double image determination unit 71 determines that a double image has been generated when the binary patterns of the five surrounding pixels all match the above-described double image pattern.

  The output selection unit 73 is supplied with the motion adaptive interpolation signal and the C signal delayed by two lines by the two-line delay unit 72 and the pull-down detection result. When the output selection unit 73 receives that the double image has been generated from the peripheral comprehensive double image determination unit 71, the output selection unit 73 moves the pixel of the C signal (field interpolation signal) to the same position in the motion adaptive interpolation signal. To avoid occurrence of a double image error.

  As described above, according to the progressive conversion unit 11 in the present embodiment, an input in which a normal signal of 60 fields / second is inserted into a signal converted to interlace by 3-2 pulldown or 2-2 pulldown. When converting a signal into a progressive signal, a field interpolation signal and a motion adaptive interpolation signal are generated, and a double image error is detected in the field interpolation signal in units of pixels to detect a double image error. In such a case, the field interpolation signal is replaced with a motion adaptive interpolation signal in units of pixels, thereby avoiding a double image error that greatly impairs the image quality, and there is no deterioration in image quality in both the insertion portion and the other portions. It becomes possible to convert to progressive well.

  It should be noted that the present invention is not limited to the above-described embodiments, and various modifications can be made without departing from the scope of the present invention.

  For example, in the above-described embodiment, double image error detection is performed using information for five vertical lines and five horizontal pixels. However, the present invention is not limited to this, and the number of lines and the number of pixels are not limited. It may be increased or decreased.

  In the above-described embodiment, the case where the frame rate is 60 fields / second has been described. However, the present invention is not limited to this, and various types of PAL signals displayed at 50 fields / second are used. It can be widely applied to display of a desired frame rate based on the format.

  In addition to the 3-2 pulldown and the 2-2 pulldown, the present invention can be widely applied to signals obtained by converting the frame rate by arranging the original image screen based on a predetermined sequence.

  In the above-described embodiment, the example in which the field interpolation signal is replaced with the motion applied interpolation signal when the pull-down detection unit 25 detects the double image portion has been described. However, the present invention is limited to this. However, the present invention can be widely applied to a case where the signal is replaced with another signal that improves the double image, for example, simply replaced with an inter-field interpolated signal.

  According to the present invention described above, since a field interpolation signal and a motion adaptive interpolation signal can be selected in units of pixels, conversion converted to interlace by 3-2 pulldown, 2-2 pulldown, or the like. Even when a normal interlace signal of, for example, 60 fields / second is inserted into the signal, both the inserted portion and the other portion can be converted into progressive well without deterioration in image quality.

It is a figure which shows an example of schematic structure of the image processing apparatus in this Embodiment. It is a figure which shows an example of an internal structure of the progressive conversion part in the image processing apparatus. It is a figure which shows an example of the internal structure of the pull-down error detection part in the progressive conversion part. It is a figure which shows the position of each signal by which line delay was carried out. It is a figure which shows the other example of the internal structure of the pull-down error detection part in the progressive conversion part. It is a figure which shows the example of the binary pattern for 5 lines. FIG. 4A is a diagram showing various scanning line structures, FIG. 2A shows an interlace signal, FIG. 2B shows a progressive signal, and FIG. 2C shows an interlace signal converted into a progressive signal by scanning line interpolation. The converted signal is shown. It is a figure explaining inter-field interpolation and intra-field interpolation. It is a figure for demonstrating 3-2 pulldown. It is a figure for explaining 2-2 pulldown. It is a figure for demonstrating the problem at the time of converting the input signal by which the image part of 60 fields / second was inserted into the signal converted into the interlace by 2-2 pulldown into a progressive signal.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 10 Front image processing part, 11 Progressive conversion part, 12 Display apparatus drive circuit, 13 Display apparatus, 20, 21 Field delay device, 22 Motion detection part, 23 Memory, 24 Motion adaptive interpolation data generation part, 31-33 Line delay , 34-37 difference absolute value calculation unit, 38 1-line difference absolute value average calculation unit, 39-42 flip-flop, 43 peripheral average calculation unit, 44-46 difference absolute value calculation unit, 47 2-line difference absolute value average calculation 48-51 flip-flop, 52 peripheral average calculation unit, 53 double image determination unit, 54 2-line delay unit, 55 output selection unit, 60-63 line delay unit, 64 average value calculation unit, 65 binarization unit 66 double image determination unit, 67-70 flip-flop, 71 peripheral total double image determination unit, 72 2-line delay unit, 73 Output selection unit, 100 image processing apparatus

Claims (12)

The original image screen is arranged based on a predetermined sequence, and the converted video signal is converted to interlace by 3-2 pulldown or 2-2 pulldown so as to match the frame rate of the input video signal. An image processing apparatus that converts an interlace signal included as a progressive signal,
A scan line located behind the current field in time and located at the same position as the interpolated scan line of the current field, or located before the current field in time and located at the same position as the interpolated scan line. Field interpolation signal generation means for generating a progressive field interpolation signal from the scanning line by field interpolation;
A double image determination means for determining whether or not the pixel of interest in the field interpolation signal is a pixel constituting a double image portion of the field interpolation signal;
A scanning line located above and below the interpolated scanning line in the current field by intra-field interpolation or temporally behind the current field and located at the same position as the interpolated scanning line; and Interpolation signal generating means for generating a progressive interpolation signal by inter-field interpolation from a scan line located before the current field in time and located at the same position as the interpolated scan line;
The double image determination unit is configured to determine, for the target pixel, a scan line that is one scan line away from a correlation between pixels of adjacent scan lines among pixels on a straight line that includes the target pixel and is orthogonal to the scan line. When the correlation between the pixels is larger, it is determined that the pixel constitutes the double image portion, and a field interpolation signal corresponding to the target pixel generated by the field interpolation signal generation unit is obtained. An image processing apparatus that replaces the interpolation signal generated by the intra-field interpolation or inter-field interpolation corresponding to the target pixel. The interpolation signal generating means for generating an interpolation signal by intra-field interpolation or inter-field interpolation has motion detection means for detecting temporal motion of the image in the current field, and the motion is detected by the motion detection means. 2. The image processing according to claim 1, wherein the interpolation signal is generated by the intra-field interpolation and the interpolation signal is generated by the inter-field interpolation when no motion is detected by the motion detecting means. apparatus.   The double image determination unit obtains the correlation for a plurality of consecutive pixels on the same scanning line including the target pixel, and determines whether the target pixel is a pixel constituting the double image portion. Item 6. The image processing apparatus according to Item 1.   The double image determining means binarizes pixels on a straight line orthogonal to the scanning line including the pixel of interest in a plurality of continuous scanning lines including the scanning line to which the pixel of interest belongs in the field interpolation signal. The image processing apparatus according to claim 1, wherein a classified binary pattern is obtained, and it is determined based on the binary pattern whether the pixel of interest is a pixel constituting the double image portion. The double image determination means, the pixels on a straight line orthogonal to the scanning line including the target pixel, and classified into two values by magnitude relation between the average value, according to claim 4, wherein finding the binary pattern image Processing equipment. The double image determining means obtains the binary pattern for a plurality of consecutive pixels on the same scanning line including the target pixel, and determines whether the target pixel is a pixel constituting the double image portion. The image processing apparatus according to claim 4 . The original image screen is arranged based on a predetermined sequence, and the converted video signal is converted to interlace by 3-2 pulldown or 2-2 pulldown so as to match the frame rate of the input video signal. An image processing method for progressively converting an interlace signal included as
A scan line located behind the current field in time and located at the same position as the interpolated scan line of the current field, or located before the current field in time and located at the same position as the interpolated scan line. A field interpolation signal generation step for generating a progressive field interpolation signal from the scanning line by field interpolation;
A double image determination step of determining, on a pixel-by-pixel basis, whether or not the pixel of interest in the field interpolation signal is a pixel constituting a double image portion of the field interpolation signal;
A scanning line located above and below the interpolated scanning line in the current field by intra-field interpolation or temporally behind the current field and located at the same position as the interpolated scanning line; and An interpolation signal generating step for generating a progressive interpolation signal by inter-field interpolation from a scanning line located before the current field in time and located at the same position as the interpolated scanning line;
In the double image determining step, for the target pixel, among the pixels on the straight line that includes the target pixel and is orthogonal to the scan line, the scan line that is one scan line away from the correlation between the pixels of the adjacent scan lines When the correlation between the pixels is larger, it is determined that the pixel constitutes the double image portion, and the field interpolation signal corresponding to the target pixel generated by the field interpolation signal generation step is An image processing method for replacing with the interpolation signal generated by the intra-field interpolation or inter-field interpolation corresponding to the target pixel. The interpolation signal generation step of generating an interpolation signal by intra-field interpolation or inter-field interpolation has a motion detection step of detecting temporal motion of the current field image, and motion is detected in the motion detection step. 8. The image according to claim 7 , wherein the interpolated signal is generated by the intra-field interpolation when the motion is detected, and the interpolated signal is generated by the inter-field interpolation when no motion is detected by the motion detecting means. Processing method. In the double image determination step, the correlation is obtained for a plurality of consecutive pixels on the same scanning line including the target pixel, and it is determined whether or not the target pixel is a pixel constituting the double image portion. Item 8. The image processing method according to Item 7 . In the double image determining step, pixels on a straight line orthogonal to the scanning line including the target pixel in a plurality of continuous scanning lines including the scanning line to which the target pixel in the field interpolation signal belongs are binarized. 8. The image processing method according to claim 7 , wherein a classified binary pattern is obtained, and whether or not the pixel of interest is a pixel constituting the double image portion is determined based on the binary pattern. In the double image determination step, the pixels on a straight line orthogonal to the scanning line including the target pixel, and classified into two values by magnitude relation between the average value, according to claim 10, wherein finding the binary pattern image Processing method. In the double image determining step, the binary pattern is obtained for a plurality of consecutive pixels on the same scanning line including the target pixel, and it is determined whether the target pixel is a pixel constituting the double image portion. The image processing method according to claim 10 .

Images