JP2012182691A - Image conversion device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To solve such problems of 3:2 pulldown used when displaying a video signal of movie picture of 24 Hz by 60 Hz that it causes unnatural movement because the video signal is displayed at non-uniform intervals for the 24 Hz, and that the movie mood is lost in film de-judder when a conventional frame rate is used.SOLUTION: When a video signal of 24 Hz is displayed by 60 Hz, a motion vector is detected from the video of 24 Hz. In order to equalize the deviation width from the moving trajectory of an object (=the moving trajectory of sight line) estimated from the motion vector, an interpolation frame is generated using a motion vector detected in some frames exceeding the deviation width, out of the frames used in the 3:2 pulldown. Consequently, a display not loosing the movie mood can be presented while reducing unnatural movement of the 3:2 pulldown.

Description

  The present invention detects a motion vector between frames of a video signal, and generates an interpolation frame using the detected motion vector, and in particular, displays a movie material shot at a frame frequency of 24 Hz at 60 Hz. The present invention relates to an image conversion apparatus.

  With the progress of video signal processing technology and semiconductor technology, it has become possible to detect a motion vector from a video signal and convert the frame frequency of the video signal using the detected motion vector. Such frame frequency conversion is performed, for example, when converting a video signal shot in the PAL system (50 Hz) to the NTSC system (60 Hz) or converting a movie (24 Hz) to a television signal (60 Hz or 50 Hz). Can be used for

  In general, movies are shot at a frame rate of 24 Hz. When this video is displayed in the NTSC format (60 Hz), frame rate conversion (telecine conversion) by the 3: 2 pull-down method has been used. This 3: 2 pull-down displays one image of a movie alternately in 3 frames and 2 frames of 60 Hz (FIG. 12). As a result, the image displayed at 60 Hz is displayed the same image a plurality of times, and is displayed at a time interval different from the original frame period, resulting in motion blur (film judder), Unnaturalness will occur in the way of movement.

  On the other hand, in Patent Document 1, a motion vector is detected from a 24 Hz video, and an interpolation frame is generated and displayed in accordance with a display timing of 60 Hz using the motion vector, thereby causing smoothness without unnaturalness. It is disclosed that it is possible to display various movements. Such frame rate conversion is called a film dejadder. However, when 24 Hz video is converted to 60 Hz and displayed using such a motion vector, the motion is smooth, but on the other hand, it may have a different impression from a movie shot at 24 Hz. Will be lost. In addition, since there is a large difference in the smoothness of motion depending on whether a correct motion vector is detected, when the correct motion vector cannot be detected due to a large motion, a large image quality deterioration occurs.

  On the other hand, in Patent Document 2, the degree of film judder reduction is achieved by bringing the position on the time axis for generating the interpolation frame closer to the input 3: 2 pull-down timing than the timing of equally dividing 24 Hz. It is disclosed that can be adjusted.

  The calculation of the motion vector used for generating the interpolation frame as in Patent Document 1 and Patent Document 2 described above is usually performed by dividing an image of one frame into a plurality of blocks each having a predetermined number of pixels (M × N pixels). The detection is performed by a method such as searching for a position where the correlation is greatest between consecutive frames for each block. The position difference at this time corresponds to a motion vector for the block. For example, as shown in FIG. 13, when the motion vector of the target block (1301) in the frame (F) is detected, a position having the largest correlation with this block is searched in the frame (F-1). Is detected as a motion vector (1302).

  As shown in FIG. 14, the generation of the interpolation frame using the motion vector can be performed by moving at least one pixel or pixel block of the frame before and after the generated interpolation frame along the motion vector. . At this time, the position on the time axis for generating the interpolation frame (hereinafter referred to as the interpolation phase) can be arbitrarily selected between the frame (F-1) and the frame (F). In FIG. In this example, the interpolation phase is -1) to 1/3 of the interpolation phase. As described above, it is possible to convert the frame frequency by detecting a motion vector from the video signal and generating an interpolation frame using the detected motion vector.

  The method of using a motion vector in frame frequency conversion is to detect by comparing frames in which motion vectors are continuous. Therefore, the movement of an object can be detected correctly, but movement such as rotation and enlargement / reduction is possible. May not be detected correctly. In addition, a correct motion vector may not be detected for a region that is hidden in the background of a moving object, a region that appears from the background, or a region that is included in only one of successive frames, such as deformation of the object. In addition, normal motion vectors are often detected by searching a predetermined range based on the target block to be detected, and when a motion exceeding this search range cannot be detected There is.

  When converting the frame frequency using a motion vector, if the detected motion vector is not accurate, it is called a halo around a moving object or the like in an interpolated frame and a video in which these are continuous. It is known that noise occurs. This halo is due to the generation of an inaccurate interpolation frame. Therefore, when the ratio of the interpolation frame is large in the displayed frame or when the interpolation frame is displayed for a long time, the halo is prominent. There is a case.

  For example, when a 60 Hz video is converted to 120 Hz, one of the two frames is an interpolation frame, and the time for which the interpolation frame is displayed is 1/120 second, so halo is relatively difficult to recognize. On the other hand, when a 24 Hz video is converted to 60 Hz and a film dejudger is realized, assuming that the 24 Hz video is (1), (2), (3), the 60 Hz video is (1), (1.4 ), (1.8), (2.2), (2.6), (3) must be output. Here, (1.4), (1.8), (2.2), and (2.6) are generated interpolation frames, and four of the five frames are generated. One interpolation frame is displayed for 1/60 second. For this reason, the film dejadder has a great influence on image quality degradation when the motion vector is wrong.

  Patent Document 2 also discloses the effect of reducing this halo by bringing the position on the time axis for generating the interpolation frame closer to the input. This is because the movement amount from the input frame is reduced when the same vector is used as the interpolation frame generation position approaches the input frame, and the influence of the wrong motion vector is reduced.

JP-A-9-172618 International Publication No. 2008/102826

  Patent Document 2 discloses a generation position of an interpolation frame when a 24 Hz movie material is displayed at 120 Hz, that is, when the frame rate is an integral multiple. On the other hand, when 24 Hz is converted to 60 Hz, the magnification is 2.5 times. Therefore, even if this method is applied as it is, a certain effect can be obtained, but the movement caused by the non-uniformity of 3: 2 pull-down. The unnaturalness still remains. In addition, even if the generation position of the interpolation frame is brought close to the input frame, it remains the same that four of the five displayed frames are interpolation-generated frames, and therefore the effect of reducing the influence of the wrong motion vector vector Is not enough.

  An object of the invention of the present application is to suppress deterioration in image quality due to a motion vector, as compared with the prior art, when performing frequency conversion of a video signal.

  The image conversion apparatus according to the present application is an image conversion apparatus that converts a first video signal having a first frame rate into a second video signal having a second frame rate higher than the first frame rate. A motion vector detection unit that detects a motion vector from at least two temporally continuous frames of the first video signal, an interpolation frame based on the first video signal, the motion vector, and the interpolation phase. A frame generation unit that generates and generates a second video signal from the frame of the first video signal and the interpolation frame; and the interpolation frame and the motion vector are used to generate the interpolation frame. The magnitude of the phase difference from the movement trajectory is greater than a predetermined magnitude between the frame of the first video signal included in the second video signal and the movement trajectory. Comprising the magnitude of the difference, and a interpolation control unit that determines the interpolation phase of the interpolation frame so as to be substantially identical.

  As a result, when a phase difference between the movement trajectory and the frame constituting the second video signal occurs, the magnitude of the phase difference becomes substantially uniform, so that an object included in the viewed video is blurred ( A phenomenon in which a plurality of objects are displayed in an overlapping manner can be suppressed as compared with the conventional case.

  In the image conversion apparatus, the interpolation control unit may determine that the phase difference between the interpolation frame and the movement locus indicated by the motion vector is a frame of the first video signal included in the second video signal. The interpolation phase of the interpolation frame may be determined so as to be substantially the same as the smallest phase difference having a magnitude greater than or equal to a predetermined value between the movement trajectory and the movement trajectory.

  Thereby, the phase difference between the movement trajectory and the frame included in the second video signal can be reduced as a whole, and an image with less blur can be configured.

  In the image conversion apparatus, the first video signal may be a video signal having a frame rate of 24 Hz, and the second video signal may be a video signal having a frame rate of 60 Hz.

  Thereby, it is also possible to convert a video image such as a movie film into a TV signal such as NTSC.

  In the image conversion apparatus, the interpolation control unit periodically generates a phase difference between the movement locus indicated by the motion vector and the frame of the generated second video signal, and the second video An interpolation phase of the interpolation frame is determined so that a phase difference of a frame having a phase difference greater than or equal to a predetermined value among the signal frames is substantially constant, and the frame generation unit is configured to The second video signal may be generated so that the ratio of the video frame to the interpolation frame is 3 to 2.

  Thereby, even when blurring cannot be completely suppressed, an image having characteristics like film judder can be generated. In addition, the generated second video signal has a composition ratio of the frame of the first video signal and the interpolated frame of 3: 2, which is more reliable than the conventional 1: 4. The video conversion device can generate the video.

  Further, in the image conversion apparatus, the interpolation control unit sets the magnitude of the phase difference between the interpolation frame and the frame of the first video signal closest to the interpolation frame to approximately 1/5. May be.

  As a result, the interpolation frame included in the second video signal has a phase difference of 0.2 (1/5) from the original frame (the frame of the first video) having the smallest phase difference. When generating an interpolation frame, it is possible to generate a more reliable frame.

  As described above, according to the image conversion apparatus of the present invention, it is possible to suppress degradation in image quality due to a motion vector when performing frequency conversion of a video signal as compared with the conventional art. In particular, in the case of converting a 24 Hz video to a 60 Hz video, it is possible to reduce image quality degradation due to an incorrect motion vector by reducing the ratio of generated interpolation frames and reducing the amount of movement by vectors. Further, in the case of this conversion, it is possible to leave a movie-like feeling and improve the unnatural movement of 3: 2 pull-down.

The figure which shows the structure of the image converter in embodiment of this invention The figure which shows the timing relationship of the production | generation of the input image and interpolation frame of film dejudder processing in embodiment of this invention Diagram showing video displayed using 3: 2 pull-down Diagram showing how the 3: 2 pull-down video looks Diagram showing film dejada processing and visual appearance Diagram showing how the 2: 2 pull-down video looks Diagram showing the amount of deviation from the line of sight in 2: 2 pull-down video The figure which shows the deviation | shift amount from the eyes | visual_axis in the image | video pulled down 3: 2 The figure which shows the timing relationship of the production | generation of an input image | video and an interpolation frame in embodiment of this invention The figure which shows the production | generation method of the interpolation frame in embodiment of this invention The figure which shows the deviation | shift amount from the eyes | visual_axis at the time of performing frame rate conversion in embodiment of this invention Diagram showing 3: 2 pull-down of the prior art Diagram explaining the mechanism of motion vectors The figure which shows the example of the production | generation of the interpolation frame from a motion vector

(Embodiment 1)
FIG. 1 is a block diagram illustrating a main configuration of an image conversion apparatus 201 according to an embodiment of the present invention. The image conversion apparatus 201 includes a video memory 202, a vector detection circuit 203, an interpolation control circuit 204, a vector memory 205, and a frame generation circuit 206. Note that it is also possible to realize the invention described in this embodiment mode by using the image conversion apparatus 201 to which the display unit 207 is added as an image display apparatus. In the following description, the image conversion apparatus will be described for the sake of simplicity.

  The image conversion apparatus 201 detects a motion vector from the input video signal 208, generates an interpolation frame using the motion vector, and displays the output video signal 215. Here, the input video signal 207 is a video signal of 24 Hz, and the output video signal 215 is a video signal of 60 Hz. The display unit 207 displays the 60 Hz video signal. That is, the image conversion apparatus 201 receives a video signal of 24 Hz, converts the frame frequency, and displays a video at 60 Hz. The display unit 207 is not particularly limited as long as it can display a video signal such as an LCD display or a PDP display.

  First, a case of performing a general frame rate conversion (film dejada) from 24 Hz to 60 Hz according to the present embodiment will be described. A video signal 208 input to the image conversion apparatus 201 is input to the video memory 202 and the vector detection circuit 203.

  The video memory 202 is a memory that can store an input video signal for at least three frames and can read any stored frame. The video memory 202 stores the input video signal, reads out the video signal one frame before the input video signal 208, and outputs it to the vector detection circuit 203.

  The vector detection circuit 203 divides the input video signal 208 into blocks of 8 pixels × 8 pixels, and searches for the position with the highest correlation from the previous frame video signal 209 input from the video memory 202 for each block. To detect a motion vector. The search range at this time is a range of, for example, horizontal ± 64 pixels and vertical ± 32 lines with reference to a block for detecting a motion vector, and the position with the largest correlation is obtained in this range. Also, as the correlation value, the absolute value of the difference between the value of each pixel included in the block and the value of the pixel at the position to be compared is summed up for the entire block (SAD: Sum of Absolute Difference, sum of absolute differences) Can be used. The size of the block is not limited to this, and it may be smaller or larger than this. It is also possible to use methods other than SAD as correlation values, and as a search method, many methods for reducing the amount of processing and detecting motion vectors efficiently are known. Is also possible. The vector detection circuit 203 outputs the detected motion vector 210 detected from the input video signal 208 and the previous frame video signal 209 to the vector memory 205.

The interpolation control circuit 204 determines the following control contents. In particular,
(1) Which motion vector is read out of motion vectors corresponding to two frames stored in the vector memory 205 described later,
(2) Which two frames are read out as frames before and after an interpolation frame generated from a plurality of frames of video signals stored in the video memory 202,
(3) In which interpolation phase between two frames read out from the video memory 202, an interpolation frame is generated,
These are the control contents. The interpolation control circuit 204 outputs control signals based on the determined control contents to the video memory 202, the vector memory 205, and the frame generation circuit 206, respectively. Details of the interpolation control circuit 204 will be described later.

  The interpolation control circuit 204 inputs a frame selection signal 213 for determining two frames used for interpolation to the video memory 202. The video memory 202 outputs the two frames designated by the frame selection signal 213 to the frame generation circuit 206 as the previous and next frame video signals 214.

  The vector memory 205 is a memory for storing a motion vector detected by the vector detection circuit 203, and is for absorbing a time difference between writing from the vector detection circuit 203 and reading from a frame generation circuit 206 described later. The vector memory 205 only needs to have a capacity corresponding to this time difference, but it is assumed here that a vector corresponding to two frames of the input video can be stored.

  When a vector selection signal 210 for selecting a vector to be used for interpolation is input from the interpolation control circuit 204, the vector memory 205 outputs the vector specified by the vector selection signal 210 to the frame generation circuit 206 as an interpolation motion vector 211. To do.

  Details of the control contents of the interpolation control circuit 204 will be described below. FIG. 2 shows an input video signal 301 (208), a previous frame video signal 302 (209), a detected motion vector 303 (210), an interpolation motion vector 304 (212), a previous frame 305 and a rear frame 306 of the preceding and following frame video signals. (214) and the timing relationship of the interpolation phase 307 (interpolation phase control signal 215) when performing film dejudder.

  The vector detection circuit 203 detects a motion vector at a cycle of 24 Hz of the input video. The vector detection circuit 203 calculates a detected motion vector 303 between the input video signal 301 and the previous frame video signal 302. For example, the vector detection circuit 203 detects a motion vector from the frame (0) input as the previous frame video signal 302 while the frame (1) is input as the input video signal 301, and calculates the calculated motion. Write the vector to the vector memory 205. Similarly, the vector detection circuit 203 detects a motion vector between the input frame and the previous frame, and writes the result in the vector memory 205.

  Since the interpolated video signal is performed at 60 Hz, which is the frequency of the video to be output, the output of each control signal from the interpolation control circuit 204 is also output at a period of 60 Hz. In the case of film dejudger, the input video signal is frame (0), frame (1), frame (2), frame (3), frame (4), frame (5), and if this is the reference, output The video signal to be performed is frame (0), frame (0.4), frame (0.8), frame (1.2), frame (1.6), frame (2), frame (2.4), frame (2.8), frame (3.2), frame (3.6), and frame (4). Therefore, the interpolation control circuit 204 selects a motion vector detected between the frame (1) and the frame (0) as the interpolation motion vector 304 at the timing when the output video becomes the frame (0.4). Is output as the vector selection signal 211. Further, the interpolation control circuit 204 outputs a frame selection signal 213 for outputting the frame (0) to the previous frame 305 of the preceding and following frame video signal, the frame (1) to the subsequent frame 306 of the preceding and following frame video signal, and the frame generation circuit 206. And 0.4 is output to the frame generation circuit 206 as the interpolation phase 307.

  At the timing when the output video becomes the frame (0.8), the interpolation control circuit 204 uses a signal for selecting a motion vector detected between the frame (1) and the frame (0) as the motion vector 304 for interpolation. Output as a selection signal 211, and output a frame selection signal 213 for outputting frame (0) to the previous frame 305 of the preceding and following frame video signal, frame (1) to the subsequent frame 306 of the preceding and following frame video signal, and interpolation 0.8 is output as the phase 307.

  At the timing when the output video becomes the frame (1.2), the interpolation control circuit 204 uses a signal for selecting the motion vector detected between the frame (2) and the frame (1) as the interpolation motion vector 304. Output as a selection signal 211, and output a frame selection signal 213 for outputting frame (1) to the previous frame 305 of the preceding and following frame video signal, frame (2) to the subsequent frame 306 of the preceding and following frame video signal, and interpolation 0.2 is output as the phase 307.

  At the timing when the output video becomes the frame (1.6), the interpolation control circuit 204 uses a signal for selecting a motion vector detected between the frames (2) and (1) as the interpolation motion vector 212. The selection signal 211 is output, and the frame selection signal 213 for outputting the frame (1) to the previous frame 305 of the preceding and following frame video signal and the frame (2) to the subsequent frame 306 of the preceding and following frame video signal is output. As 307, 0.6 is output.

  Interpolation is not necessary at the timing when the output video becomes frame (0), so the motion vector 212 for interpolation is not necessary, and the front frame 305 and the rear frame 306 of the preceding and following frame video signals need only be the frame (0). 0 is output as the phase 307. Thereafter, such an output video is repeated for a period of 5 frames.

  The interpolation control circuit 204 appropriately selects an input frame and a motion vector necessary for the interpolation frame generated as described above, and outputs a control signal for inputting these to the frame generation circuit 206. In accordance with this, the interpolation phase 307 is output to the frame generation circuit 206.

The frame generation circuit 206 is designated by the interpolation phase 307 using the previous frame 305 and the subsequent frame 306 input as the previous and subsequent frame video signals, and the interpolation motion vector 304 corresponding to the motion between the two frames. Generate and output an interpolated frame at the phase position.
As described above (FIG. 14), the interpolation frame can be generated by moving the pixel of the previous frame or the pixel of the subsequent frame along the interpolation motion vector. At this time, it is also possible to generate an interpolation frame using pixels moved from either one, such as using only a frame closer to the interpolation phase, and a certain percentage of pixels moved from both frames, or It is also possible to generate by mixing at a rate corresponding to the interpolation phase.

  The output video signal 216 generated by the frame generation circuit 206 as described above is displayed on the display unit 207. Since the 24 Hz input video signal 301 is converted into a 60 Hz video signal by the above processing and displayed on the display unit 207, it is displayed as a smooth video without film judder.

  Next, the cause of film judder when 24 Hz video is displayed with 3: 2 pull-down will be described. FIG. 3 shows an example in which a scene where a ball crosses the screen is photographed at 24 Hz and displayed at 60 Hz using 3: 2 pull-down. In the displayed 60 Hz video, frames taken at 24 Hz are alternately displayed in 3 frames and 2 frames. FIG. 4 is a graph showing the relationship between the ball display position and time in the 60 Hz image.

  It is known that the viewer's line of sight moves so as to follow a moving object when a person sees something that moves almost uniformly as in this example. In the example of FIG. 4, the viewer's line of sight follows the displayed ball and moves like a movement track of the line of sight in the figure. Here, in the frames 2 and 7, the display position of the ball coincides with the movement track of the line of sight, but in other frames, the ball is displayed at a position shifted from the movement track of the line of sight. Frames 1, 4, and 6 appear to be behind the position of the ball that the line of sight is chasing, and frames 3, 5, and 8 appear to have balls on the front. As a result, the balls are moving uniformly. Appears blurry back and forth. That is, it becomes a state in which a plurality of one object appears to overlap. Such a state is called film judder.

  The frame rate conversion from 24 Hz to 60 Hz, which has the effect of film dejudder to suppress the above-mentioned film judder, generates an interpolated frame in which the ball is moved to a position that coincides with the movement locus of the line of sight as shown in FIG. Is equivalent to As a result, it can be seen as a smooth movement in which film judder is suppressed.

  On the other hand, in a movie theater, a 24 Hz video is displayed at 48 Hz, and a display corresponding to 2: 2 pull-down is performed. The appearance in this case is as shown in FIG. 6, and film judder occurs. As shown in FIG. 6, the film judder generated in the movie theater display is shifted evenly and alternately with respect to the movement locus of the line of sight, and this state often feels like a movie. On the other hand, in the frame rate conversion from 24 Hz to 60 Hz by the film dejadder as described above, since such a film judder is completely eliminated, the quality of a movie is lost.

  On the other hand, the film judder generated by 3: 2 pull-down has a non-uniform displacement with respect to the movement track of the line of sight in a cycle of 5 frames, so it looks different from 2: 2 pull-down, and the movement is jerky. I feel it. FIG. 7 shows a change in how the line of sight moves from the locus of movement in the case of 2: 2 pulldown, and FIG. 8 shows a change in how the line of sight moves from the track of movement in the case of 3: 2 pulldown. As described above, in the 3: 2 pull-down, the magnitude of the deviation from the movement locus of the line of sight is not uniform, so that it feels like a jerky and unnatural movement.

  On the other hand, the present embodiment discloses a display method that leaves film judder close to a movie theater in 60 Hz display and reduces nonuniformity caused by 3: 2 pull-down. Hereinafter, processing contents of the interpolation control circuit 204 of the image conversion apparatus 201 will be described.

  As shown in FIG. 9, the interpolation control circuit 204 outputs a frame selection signal 213, a detected motion vector 1003 (210), and an interpolation phase 1007 (interpolation phase control signal 215) at a cycle of 5 frames as follows.

  1) The interpolation control circuit 204 outputs a frame selection signal 213 for outputting the frame (0) as the previous frame 1005 of the preceding and following frame video signals to the video memory 202, and outputs 0 as the interpolation phase 1007 to the frame generation circuit 206. To do. In this step, since no interpolation frame generation is required, the interpolation motion vector 1004 (212) is unnecessary.

  2) The interpolation control circuit 204 outputs to the video memory 202 a frame selection signal 213 for outputting the frame (0) in the previous frame 1005 and the frame (1) in the subsequent frame 1006 of the preceding and following frame video signals. The interpolation control circuit 204 outputs a vector selection signal 211 for selecting a motion vector detected between the frame (1) and the frame (0) to the vector memory 205, and sets 0.2 as the interpolation phase 1007 from the vector memory 205. The data is output to the frame generation circuit 206.

  3) The interpolation control circuit 204 outputs a frame selection signal 213 for outputting the frame (1) as the previous frame 1005 of the preceding and following frame video signals to the video memory 202, and outputs 0 as the interpolation phase 1007 to the frame generation circuit 206. To do. In this step, since no interpolation frame is required, the interpolation motion vector 1004 is unnecessary.

  4) The interpolation control circuit 204 outputs a frame selection signal 213 for outputting the frame (1) as the previous frame 1005 of the preceding and following frame video signals to the video memory 202, and sets (0) as the interpolation phase 1007 to the frame generation circuit 206. Output to. At this time, since no interpolation frame is required to be generated, the interpolation motion vector 1004 is unnecessary.

  5) The interpolation control circuit 204 outputs to the video memory 202 a frame selection signal 213 for outputting the frame (1) in the previous frame 1005 and the frame (2) in the subsequent frame 1006 of the preceding and following frame video signals. Further, the interpolation control circuit 204 outputs a vector selection signal 211 for selecting a motion vector detected between the frame (2) and the frame (1) to the vector memory 205, and sets 0.8 as the interpolation phase 1007 from the vector memory 205. The data is output to the frame generation circuit 206.

  As a result, the input video signal is frame (0), frame (1), frame (2), frame (3), frame (4), and frame (5). Is frame (0), frame (0.2), frame (1), frame (1), frame (1.8), frame (2), frame (2.2), frame (3), frame (3 ), Frame (3.8), and frame (4).

  FIG. 10 is a diagram similar to FIGS. 4 and 5 showing the state of interpolation in the above case. Also, the magnitude of the deviation between the line-of-sight movement locus and the interpolated video in this case is as shown in FIG. In the example of FIG. 11, the magnitudes of the shift (phase difference) between the movement trajectory and the video are indicated by 0 in frames 2 and 7. The other frames are fair so that the magnitude of the phase difference is substantially constant. In this way, with the configuration described in the present embodiment, since the interpolation phase is determined so that the magnitude of the deviation from the movement locus of the line of sight is uniform within 5 frames, although the film judder remains, 3 : Non-uniformity caused by 2 pull-down is reduced. That is, in a frame in which the magnitude of the phase difference between the movement trajectory and the interpolated video is greater than or equal to a predetermined value (in the case of FIG. 11, the phase difference is greater than 0), the interpolation frame and the movement trajectory included in the interpolated video are displayed. The phase of the interpolated frame is determined so that the magnitude of the phase difference between and the original video included in the interpolated video and the magnitude of the phase difference between the movement trajectories are substantially the same. As a result, the magnitude of the shift (phase difference) of the interpolated video is substantially constant, so that the video non-uniformity that is likely to occur during uneven pull-down processing is reduced. Further, as shown in FIG. 11, the deviation between the movement locus of the line of sight and the interpolated image is periodically generated and the magnitude of the deviation is similar to the case of the 2: 2 pull-down example shown in FIG. Since it is constant, it is possible to display an image with an impression close to that of a movie theater.

  Also, the fact that the magnitude of the phase difference is substantially constant does not necessarily mean that all the phase differences are completely the same. For example, in the above description, the magnitude of this phase difference is described as 0.2, but it may be in the range of 0.15 to 0.25, for example. In this case, an allowable range of 25% before and after the phase difference is provided. That is, it is sufficient if the magnitude of the phase difference is within a predetermined range before and after the target magnitude of the phase difference.

  In addition, by controlling the interpolation phase as described above, the frame rate conversion (film dejada) from 24 Hz to 60 Hz is an interpolation frame in which 4 out of 5 frames of the output frame are generated. On the other hand, in the interpolation method disclosed in the present invention, only 2 frames out of 5 frames are generated. As described above, the ratio of the interpolation frame generated using the motion vector included in the output video affects the magnitude of image quality degradation when an incorrect motion vector is detected. With this interpolation method, it is possible to perform frame rate conversion with less image quality degradation than in the prior art. At this time, since the number of interpolation frames to be generated is half of the number of frames to be displayed, the processing amount necessary for generating the interpolation frames can be reduced to half that of the conventional film dejudger.

  Furthermore, in the interpolation method disclosed by the present invention, only 0.2 and 0.8 are used as the interpolation phase of the generated interpolation frame. As described above, when the interpolation phase for generating a frame approaches the input frame, the amount of movement from the input frame decreases, so that the influence of an incorrect motion vector decreases. For this reason, in the interpolation method disclosed by the present invention, the influence of the wrong motion vector on the image quality is small as compared with the conventional film dejudger using the interpolation phases of 0.4 and 0.6.

  As described above, if the configuration of the present embodiment is used, the interpolation phase is smaller and the interpolation phase close to the input frame can be used compared to the conventional case. Small image quality degradation.

  As described above, according to the interpolation method of the present invention, the non-uniformity caused by 3: 2 pull-down is reduced, the film judder that is close to a movie theater is left, and the motion vector is erroneously detected. It is possible to convert the frame rate from 24 Hz to 60 Hz with small image quality degradation.

  In the above embodiment, the description is based on the assumption that 24 Hz video is input as the input video signal. However, even if the input video signal is a 60 Hz video signal that is pulled down 3: 2. It doesn't matter. If a 24 Hz video signal before being 3: 2 pulled down can be appropriately selected from a 3: 2 pulled down 60 Hz video signal, the same processing can be performed.

  Also, the timing relationship between the input video signal 1001 and the processing of each block and the signal shown in FIG. 9 is an example, and depending on the capacity of the video memory 202 and the vector memory 205, the processing can be performed at a different timing. .

  In the present embodiment, the hardware configuration shown in FIG. 1 has been described as an example, but the present invention is not limited to this mode. The invention described in this embodiment can also be realized by software or the like equivalent to the configuration shown in FIG.

  The present invention can be used for an apparatus that detects a motion vector from a video signal, converts the frame rate using the detected motion vector, and displays the same.

DESCRIPTION OF SYMBOLS 201 Image converter 202 Video memory 203 Vector detection circuit 204 Interpolation control circuit 205 Vector memory 206 Frame generation circuit 207 Display unit 208 Input video signal 209 Previous frame video signal 210 Detection motion vector 211 Vector selection signal 212 Motion vector for interpolation 213 Frame selection Signal 214 Front and rear frame video signal 215 Interpolation phase control signal 216 Output video signal 301, 1001 Input video signal 302, 1002 Previous frame video signal 303, 1003 Detected motion vector 304, 1004 Interpolation motion vector 305, 1005 Previous frame video signal Frames 306 and 1006 Frames before and after the frame image signal 307 and 1007 Interpolation phase

Claims (5)

In an image conversion device for converting a first video signal having a first frame rate into a second video signal having a second frame rate higher than the first frame rate,
A motion vector detection unit for detecting a motion vector from at least two temporally continuous frames of the first video signal;
A frame for generating an interpolation frame based on the first video signal, the motion vector, and an interpolation phase, and generating a second video signal from the frame of the first video signal and the interpolation frame A generator,
When generating the interpolation frame, the magnitude of the phase difference between the interpolation frame and the movement trajectory indicated by the motion vector is determined based on the first video signal frame and the movement trajectory included in the second video signal. An interpolation control unit for determining an interpolation phase of the interpolation frame so as to be substantially the same as the magnitude of the phase difference having a predetermined size or more, and
An image conversion apparatus comprising: The interpolation control unit is configured such that the magnitude of the phase difference between the interpolation frame and the movement locus indicated by the motion vector is a predetermined value between the frame of the first video signal included in the second video signal and the movement locus. Determining the interpolation phase of the interpolation frame so as to be substantially the same as the smallest of the magnitudes of the phase differences having the above-mentioned magnitudes;
The image conversion apparatus according to claim 1. The first video signal is a video signal having a frame rate of 24 Hz,
The second video signal is a video signal having a frame rate of 60 Hz.
The image conversion apparatus according to claim 1. The interpolation control unit
A phase difference between the movement trajectory indicated by the motion vector and the generated frame of the second video signal is periodically generated, and the magnitude of the phase difference greater than or equal to a predetermined value in the frame of the second video signal is determined. Determining the interpolation phase of the interpolated frame so that the magnitude of the phase difference of the frame having is substantially constant,
The frame generation unit generates the second video signal so that a ratio of a frame of the first video and an interpolation frame is 3 to 2.
The image conversion apparatus according to claim 3. The interpolation control unit sets the magnitude of the phase difference between the interpolation frame and the frame of the first video signal closest to the interpolation frame to approximately 1/5.
The image conversion apparatus according to claim 4.