JP2010161712A - Frame rate converting device and frame rate converting method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a frame rate converting device capable of detecting a correct motion vector by extending a motion search range without increasing the throughput of a motion detection processing. <P>SOLUTION: The frame rate converting device comprises: a reduction processing section 903 for reducing an image in an inputted moving image; a motion vector detecting section 904 for detecting a motion vector between detected motion vectors; an enlargement processing section 905 for enlarging the detected motion vector; and a frame generating section 906 for generating a new image from the image in the moving image using the enlarged motion vector. An image reduction scale in the reduction processing section 903 and a motion vector enlargement scale in the enlargement processing section 905 are controlled, thereby controlling the motion vector search range. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a frame rate conversion apparatus and a frame rate conversion method for converting a frame rate of a moving image.

  As a method for improving hold blur in a liquid crystal display device such as a liquid crystal television, there is a technique for increasing a frame rate of a moving image by generating a new image (an image for interpolation). This technique is called FRC.

  FIG. 1 is a diagram for explaining a conventional example of a double speed conversion process performed in a liquid crystal display device. FIG. 1 shows an example of processing for converting the frame rate of a moving image to double when a moving image having a frame rate of 60 Hz is input. The images 101, 102, and 103 in the moving image 1a are images that are continuously input in time. Then, as indicated by the positions of the objects 104, 105, and 106 on the screen, the certain object 104 moves in the screen at a constant speed in the lower right direction. The moving image 1b shows an output moving image obtained by converting the frame rate of the input moving image of the moving image 1a by a factor of two. The output images 107, 109, and 111 are the same images as the input images 101, 102, and 103, and the output images 108 and 110 are newly generated images. Objects 112 and 113 in the newly generated output images 108 and 110 detect the movement direction of the object in the input image, calculate the position of the object at an intermediate time in the input image, and move from the input image. Generated by compensation.

  FIG. 2 is a block diagram showing a configuration example of a main part of the liquid crystal display device for performing the double speed conversion processing of FIG. The liquid crystal display device 200 includes a frame memory 201, a motion vector detection unit 202, and a frame generation unit 203. The frame memory 201 is a memory for storing an input image, and supplies an image to the motion vector detection unit 202 and the frame generation unit 203 at an optimal timing by control from a timing controller (not shown). The motion vector detection unit 202 reads the images 1 and 2 that are temporally continuous from the frame memory 201 and detects the motion (motion vector) of the object between the images. The frame generation unit 203 generates and outputs motion compensated images such as the output images 108 and 110 in the moving image 1b from the image 3 supplied from the frame memory 201 and the motion vector output from the motion vector detecting unit 202. Further, the frame generation unit 203 outputs the image 3 read from the frame memory 201 as it is when the input image is an output image such as the output images 107, 109, and 111 in the moving image 1b.

  FIG. 3 is a diagram for explaining an example of motion detection processing in the motion vector detection unit in FIG. The motion vector detection unit 202 detects the motion of an object between the motion source image 301 and the motion destination image 302 (for example, motion from the object 303 to the object 304). In order to correspond to the motion of a plurality of objects, the motion source image 301 is divided into a plurality of blocks in a grid pattern, and the motion vector of each block is detected. For example, the motion vector for the block 305 of the motion source image is detected from the region of the motion search range 307 centering on the block 306 at the same position of the motion destination image. In the example of FIG. 3, a block 308 is detected as a movement destination block of the block 305, and a vector from the block 305 to the block 308 becomes a motion vector.

  With reference to FIG. 4, a double speed conversion process when a 60 Hz moving image obtained by converting a 24 Hz film material into 3-2 pull-down is input will be described. FIG. 4 is a diagram for explaining a conventional example of double speed conversion processing for 3-2 film video performed in the liquid crystal display device.

  A 60 Hz moving image 4b including images 403 to 407 is obtained by performing 3-2 pull-down conversion on the 24 Hz moving image (film) 4a including the images 401 and 402. An output moving image when the same double speed conversion processing as in FIG. 1 is applied to the moving image 4b is a moving image 4c including images 408, 409, 410,. When attention is paid to the movement of the object in the moving image 4c, the movement of the object is jerky without moving at a constant speed because the stationary state and the moving state are alternately repeated. On the other hand, it is detected that the moving image 4b is a 3-2 pull-down converted image of the moving image (film) 4a, and based on the detection result, the moving image 4b (the moving image 4b corresponding to each image of the moving image 4a). When the double-speed conversion process is performed on the image of (2), the object can be moved at a constant speed as shown in the moving image 4d. Here, for the moving image 4d, the same image repeatedly input from an input image with a frame rate of 60 Hz is discarded and returned to an image with a frame rate of 24 Hz, and four images are generated between the images 411 and 416 with the frame rate of 24 Hz. A frame rate of 120 Hz is realized by creating images 412 to 415. In this way, the process of performing motion vector detection at the original 24 Hz on a moving image of a film material, generating a frame image therebetween, and smoothing the motion of the object is called a film dejada.

  A double speed conversion process when a 60 Hz moving image obtained by 2-2 pulldown conversion of a 30 Hz material such as computer graphics is input will be described with reference to FIG. FIG. 5 is a diagram for explaining a conventional example of the double speed conversion process for the 2-2 film image performed in the liquid crystal display device.

  The 60 Hz moving image 5b including the images 503 to 506 is obtained by subjecting the 30 Hz moving image 5a including the images 501 and 502 to 2-2 pull-down conversion. The output moving image when the same double speed conversion processing as in FIG. 1 is applied to the moving image 5b is a moving image 5c including images 507, 508, 509,. Paying attention to the movement of the object in the moving image 5c, the movement of the object does not move at a constant speed in the 2-2 pull-down as well as the 2-2 pull-down in FIG. Jerky. On the other hand, it is detected that the moving image 5b is an image obtained by subjecting the moving image 5a to 2-2 pull-down conversion, and based on the detection result, the moving image 5b (an image of the moving image 5b corresponding to each image of the moving image 5a). When the double speed conversion process is performed on the object, the object can be moved at a constant speed as shown in the moving image 5d. Here, for the moving image 5d, the same image repeatedly input from an input image having a frame rate of 60 Hz is discarded and returned to an image having a frame rate of 30 Hz, and three images are generated between the images 510 and 514 having a frame rate of 30 Hz. A frame rate of 120 Hz is realized by creating the images 511 to 513.

  As a technique for controlling the motion search range, Patent Document 1 discloses a frame rate conversion device that enlarges an input image signal, obtains a motion vector from an image before enlargement, and further scales the motion vector. Yes. As another technique for controlling the motion search range, Patent Document 2 discloses that when a motion vector is not detected by the first search, the image is reduced and the second search is performed, so that the same processing amount is obtained. An image encoding device that apparently expands the search range of motion vectors is disclosed.

  In addition, when performing frame rate conversion, a technique for preventing image quality deterioration when an input image signal is an image signal subjected to pull-down conversion has also been proposed (see, for example, Patent Documents 3 to 5). When the input image signal is an image signal subjected to pull-down conversion, the image display devices described in Patent Documents 3 and 5 invalidate the motion compensation processing in the rate conversion means, and the image display device described in Patent Document 4 has a rate. The intensity of the motion compensation process in the conversion means is reduced.

JP 2007-324830 A JP 2004-140794 A JP 2008-78859 A JP 2008-79295 A JP 2008-136226 A

  However, when the film judging process described with reference to FIGS. 4 and 5 is performed, if the motion search range designed for the post-pulldown frame rate is applied to the pre-pulldown frame rate image, the correct motion is achieved. The appearance frequency of scenes where vectors cannot be detected increases. On the other hand, when the motion search range designed for the frame rate before pull-down is applied, the amount of processing required for motion detection increases significantly.

  Such a problem will be specifically described with reference to FIG. FIG. 6 is a diagram showing the amount of motion of a film image taken at a frame rate of 24 Hz and the amount of motion of a video image taken at a frame rate of 60 Hz. Film images shot at a frame rate of 24 Hz are delayed in the timing at which the images are updated compared to video images shot at a frame rate of 60 Hz by a video camera. Therefore, as shown in the video image 6a composed of each frame and the film image 6b composed of each frame (frame), when the same scene is photographed, a difference occurs in the amount of motion of the object between temporally continuous images. The amount of motion of a moving image recorded at a frame rate of 60 Hz (for example, the amount of motion from the object 601 to the object 602) and the amount of motion of a moving image recorded at a frame rate of 24 Hz (for example, the amount of motion from the object 603 to the object 604) ), The ratio is 2.5 times.

  Therefore, when performing film dejada processing as described in FIG. 4 (or FIG. 5), motion search designed for a 60 Hz frame rate after pull-down is performed on an image with a frame rate of 24 Hz (or 30 Hz). When the range is applied, the amount of motion of the object exceeds the motion search range, and the appearance frequency of a scene in which a correct motion vector cannot be detected increases. Conversely, if a motion search range is set for a frame rate of 24 Hz (or 30 Hz) before pull-down, the frequency that the amount of motion of the object exceeds the search range decreases, and a correct motion vector can be detected. The amount of processing required increases significantly.

  Further, such a problem is not limited to a moving image subjected to pull-down processing. For example, in a frame rate conversion apparatus that performs frame rate conversion processing with motion compensation for both cases where the frame rate of an input moving image is a specified value (for example, 60 Hz) and lower than that, the above-described Problems arise. Therefore, in such a frame rate conversion device, when the frame rate of the input moving image is low, it is necessary to widen the motion search range in order to accurately detect the motion vector. However, with the conventional motion vector detection technique, the processing amount increases significantly when the motion search range is expanded.

  Furthermore, in order to detect a motion vector more accurately, a frame rate conversion device that generates an image using a motion vector and converts the frame rate of the moving image regardless of the frame rate of the input moving image. There is a problem that it is necessary to widen the motion search range, and on the other hand, if it is widened, the amount of processing increases significantly.

  Since the technique described in Patent Document 1 is based on the premise that the input image signal is enlarged, a scaling circuit is required, and a processing amount increases and a memory having a large capacity needs to be prepared. . The technique described in Patent Document 2 is not a technique related to frame rate conversion, and does not change the motion search range according to the frame rate. In addition, any of the techniques described in Patent Documents 3 to 5 is a technique for preventing image quality deterioration at the time of frame rate conversion for a pull-down converted moving image, that is, a moving image in which a plurality of identical images may continue. However, it is not a technique for controlling the motion search range.

  The present invention has been made in view of the above-described actual situation, and an object of the present invention is to expand a motion search range and detect an accurate motion vector without increasing the amount of motion detection processing. Another object of the present invention is to provide a frame rate conversion apparatus and a frame rate conversion method.

  In order to solve such a problem, the first technical means of the present invention detects a motion vector between images in an input moving image and generates a new image using the motion vector, thereby In the frame rate conversion device for converting a frame rate, an image reduction means for reducing an image in the moving image, a motion vector detection means for detecting a motion vector between images reduced by the image reduction means, and the motion vector detection A motion vector enlarging means for enlarging the motion vector detected by the means, and an image generating means for generating the new image from the image in the moving image using the motion vector expanded by the motion vector enlarging means. Controlling an image reduction magnification in the image reduction means and a motion vector enlargement magnification in the motion vector enlargement means. By, in which the control means controls the search range of the motion vector.

  A second technical means includes, in the first technical means, a material detection means for detecting the material of the moving image, and an image reduction magnification and an image according to a detection result of the material of the moving image in the material detection means. The search range is controlled by controlling the magnification of the motion vector.

  According to a third technical means, in the first technical means, the search range is controlled by controlling an image reduction magnification and a motion vector enlargement magnification in accordance with a frame rate of the moving image. It is what.

  According to a fourth technical means, in the first technical means, the search is performed by controlling an image reduction magnification and a motion vector enlargement magnification according to a statistical value of a motion vector detection result in the motion vector detection means. It is characterized by controlling the range.

  In a frame rate conversion method for converting a frame rate of the moving image by detecting a motion vector between images in the input moving image and generating a new image using the motion vector. The image reduction means reduces the image in the moving image, the motion vector detection means detects the motion vector between the images reduced in the image reduction step, and the motion vector enlargement. Means for enlarging the motion vector detected in the motion vector detection step; and image generation means using the motion vector expanded in the motion vector expansion step from the image in the moving image. An image generation step for generating a new image and a control means include the image reduction step. By controlling the magnification of the motion vectors in the reduced ratio between the motion vector enlarged steps of an image in, is obtained by comprising: the step of controlling the motion vector search range, the.

  According to the present invention, an accurate motion vector can be detected by expanding the motion search range without increasing the amount of motion detection processing.

It is a figure for demonstrating the prior art example of the double speed conversion process performed in a liquid crystal display device. It is a block diagram which shows the structural example of the principal part of the liquid crystal display device for performing the double speed conversion process of FIG. It is a figure for demonstrating an example of the motion detection process in the motion vector detection part of FIG. It is a figure for demonstrating the prior art example of the double speed conversion process with respect to 3-2 film image performed in a liquid crystal display device. It is a figure for demonstrating the prior art example of the double speed conversion process with respect to the 2-2 film image | video performed in a liquid crystal display device. It is the figure which showed the amount of motion of the film image image | photographed with the frame rate of 24 Hz, and the amount of motion of the video image image | photographed with the frame rate of 60 Hz. It is a figure for demonstrating the conventional motion detection process for comparing with the motion detection process which concerns on this invention. It is a figure for demonstrating the outline | summary of the motion detection process which concerns on this invention. It is a block diagram which shows the structural example of the frame rate conversion apparatus which concerns on one Embodiment of this invention. It is a block diagram which shows the structural example of the frame rate conversion apparatus which concerns on other embodiment of this invention.

  A frame rate conversion apparatus according to the present invention is an apparatus that converts a frame rate by interpolating an image with a moving image. Hereinafter, this apparatus will be described with reference to the drawings.

  FIG. 7 is a diagram for explaining a conventional motion detection process for comparison with the motion detection process according to the present invention, and FIG. 8 is a diagram for explaining an outline of the motion detection process according to the present invention. .

  When the frame rate at the time of shooting is low as in the case of the film image 6b in FIG. 6, there is no problem if the amount of motion of the object is small and within the motion search range as described in FIG. Such a problem arises. In other words, as shown in FIG. 7, when the motion vector between the motion source image 701 and the motion destination image 702 is detected, if the object 703 is actually moved to the position indicated by the object 704, the motion of this object The amount exceeds the predefined motion search range 707. In the conventional motion detection process, since detection is performed within the motion search range, a correct motion vector cannot be detected except for allowing the processing amount to be increased and expanding the motion search range.

  More specifically, the motion vector for the block 705 of the motion source image 701 is detected from the region of the motion search range 707 centering on the block 706 at the same position of the motion destination image 702. Since the block 708 is detected as the destination block of the block and the vector from the block 705 to the block 708 must be a motion vector, the block destination is detected only within the motion search range 707. Cannot be detected.

  On the other hand, in the present invention, as shown in FIG. 8, the motion vector detection process is executed after the motion source image 801 and the motion destination image 802 are reduced. By performing such reduction before motion vector detection, the motion detection target image is reduced. Therefore, even if the motion search range 808 is centered on the same block 806 as the motion search range 707, the object 803 The motion amount from the block 805 to the block 808 corresponding to the motion from the object 804 to the object 804 is within the motion search range 808, and a correct motion vector can be detected.

  As described above, the motion detection processing of the present invention in which the motion vector detection processing is performed after reducing the image does not require changing the processing amount of the motion detection itself. That is, according to the present invention, since the ratio of the motion search range to the entire screen can be increased without changing the amount of motion detection processing, the object motion detection performance can be improved.

  A frame rate conversion apparatus capable of performing the motion detection process described with reference to FIG. 8 will be described with reference to FIG. FIG. 9 is a block diagram illustrating a configuration example of a frame rate conversion apparatus according to an embodiment of the present invention.

  The frame rate conversion apparatus illustrated in FIG. 9 includes a frame memory 901, a reduction processing unit 903, a motion vector detection unit 904, an enlargement processing unit 905, and a frame generation unit 906, and a motion vector between images in the input moving image. And a frame rate of the input moving image is converted by generating a new image (interpolation image) using the motion vector. First, an embodiment that does not include the film determination unit 902 will be described. In addition, the preferable form provided with the film determination part 902 is mentioned later.

  The frame memory 901 is a memory for storing an input moving image. The reduction processing unit 903 is an example of an image reduction unit that reduces an image in an input moving image. The reduction processing unit 903 reduces the motion source image 1 and the motion destination image 2 read from the frame memory 901, and outputs them as an image 6 and an image 7, respectively. As the pixel reduction process, a simple thinning process or a process using a reduction filter (a combination of a low-pass filter process and a thinning process) can be employed, but the present invention is not limited to this.

  The motion vector detection unit 904 is an example of a motion vector detection unit that detects a motion vector between images reduced by the reduction processing unit 903. The motion vector detection unit 904 detects the motion vector I from the motion source image 6 and the motion destination image 7. As a motion vector detection method in the motion vector detection unit 904, a method in which an image is divided into blocks and a motion vector is obtained for each block is generally used, and a “block matching method”, a “gradient method”, or the like can be applied. The motion vector detection unit 904 always performs the same processing on the input image regardless of the reduction magnification in the reduction processing unit 903 and the enlargement magnification in the enlargement processing unit 905 described later.

  The enlargement processing unit 905 is an example of a motion vector enlargement unit that enlarges the motion vector detected by the motion vector detection unit 904. The enlargement processing unit 905 enlarges the motion vector I and outputs it as a motion vector II.

  With regard to the reduction processing unit 903 and the enlargement processing unit 905, in the present invention, the image reduction magnification (coefficient ka multiplied when the image is reduced) in the reduction processing unit 903 and the magnification of the motion vector in the motion vector detection unit 904 (when enlargement is performed). The multiplication coefficient kb) is in a relationship (ka × kb = 1) that always becomes 1 when the two are multiplied. Here, the value of kb (= 1 / ka) may be any value as long as kb ≧ 1 and within the capability range of the frame rate conversion apparatus. Further, ka and kb are preferably coefficients common to the vertical direction (vertical direction of the screen) and the horizontal direction (horizontal direction of the screen), and it is preferable to perform reduction or enlargement while fixing the aspect ratio of the original image and the motion vector. Different values may be used in the horizontal direction. In this case, the vertical reduction magnification kav and the vertical enlargement magnification kbv have a relationship of kav × kbv = 1, and the horizontal reduction magnification kah and the horizontal expansion magnification kbh have a relationship of kah × kbh = 1. do it.

  The frame generation unit 906 is an example of an image generation unit that generates a new image (interpolation image) from an image in the input moving image using the motion vector II expanded by the expansion processing unit 905. The frame generation unit 906 generates and outputs a motion compensated image from the image supplied from the frame memory 901 and the motion vector II. The frame generation unit 906 outputs the image 3 read from the frame memory 901 as it is when the input image is used as the output image.

  The frame rate conversion apparatus according to the present invention controls the motion vector search range in the motion vector detection unit 904 by controlling the image reduction magnification in the reduction processing unit 903 and the motion vector enlargement magnification in the enlargement processing unit 905. (That is, the motion search range) is controlled. As described above, the processing in the motion vector detection unit 904 (including the motion search range for the input image) is left as it is, and the image input thereto is reduced to reduce the substantial amount in the motion vector detection unit 904. The motion search range (substantial motion search range with respect to the image content) is changed. Then, the enlargement processing unit 905 enlarges the motion vector reduced by the reduction amount of the image.

  As described above, according to the present invention, an accurate motion vector can be detected by expanding the motion search range without increasing the amount of motion detection processing in the motion vector detection unit 904. Further, since the frame rate is converted based on the accurate motion vector detected in this way, for example, hold blur in a liquid crystal display device can be appropriately eliminated.

  Next, a preferred embodiment of the present invention including the film determination unit 902 will be described. The film determination unit 902 is an example of a material detection unit that detects a material of an input moving image. As a main feature of the frame rate conversion apparatus according to the present embodiment, a motion search is performed by controlling the image reduction magnification and the motion vector magnification according to the detection result of the moving image material in the material detection means. Control the range.

  The details of this embodiment will be described with reference to the film determination unit 902. The film determination unit 902 extracts the image material (video image, image obtained by converting the film material from 3-2 pull-down, the film material from 2-2 pull-down from the input image 4 and the input image 5 of one frame past via the frame memory 901. Converted image).

  In the present embodiment, the material determination method is not limited. For example, the similarity between temporally continuous images is measured, and the similarity is always low between consecutive images based on the history of the similarity. It can be determined as a video image when the image is moving, and can be determined as a 3-2 pull-down or 2-2 pull-down converted image when the state of high and low similarity between successive images is measured periodically . Whether the image is a 3-2 pull-down converted image or a 2-2 pull-down converted image can also be determined by the cycle.

  The reduction processing unit 903 reduces the motion source image 1 and the motion destination image 2 read from the frame memory 901 when the film determination result is determined to be a film material, and outputs them as an image 6 and an image 7, respectively. If the film determination result is determined to be a video material, the original image 1 is output as image 6 and the target image 2 is output as image 7 without being reduced.

  The motion vector detection unit 904 detects the motion vector I from the motion source image 6 and the motion destination image 7. The motion vector detection unit 904 performs the same process regardless of the film determination result (video material, film material). The enlargement processing unit 905 enlarges the motion vector I when the film determination result is determined to be a film material, and outputs it as a motion vector II. If the film determination result is determined to be a video material, the enlargement process is not performed and the motion vector I is output as it is as the motion vector II.

  As described above, in the present embodiment, the material of the image is determined, and if the frame rate of the material is low (lower than the predetermined frame rate) according to the determination result, the input image is reduced and then motion search is performed. The motion search range (substantial motion search range) is expanded by expanding the motion vector obtained by this motion search. As a specific example, the reduction ratio is controlled only when the input moving image is smaller than the frame rate of a prescribed value (for example, 60 Hz) (for example, 24 Hz). Supplementally, even if the frame rate of the input moving image is the specified value (60 Hz), when the 2-2 pulldown is detected by the material detection, the frame rate at the time of shooting is equivalent to 30 Hz, and when the 3-2 pulldown is detected. Since the frame rate at the time of shooting is equivalent to 24 Hz, it is determined that the frame rate of the actually shot video is low, and the reduction rate of the image is controlled according to the frame rate of the material, and the substantial motion search range is set. Expanding. The enlargement of the motion vector is executed at the enlargement factor corresponding to the reduction factor as the compensation.

  Also, in the present invention, the target motion image is reduced by reducing the target image, so that the substantial motion search range is enlarged, and the accuracy of the detected motion vector is poor if the enlargement is too large. turn into. Therefore, as illustrated, such control is performed only when the frame rate of the captured video is low, and when it is high (when the moving distance of an object between adjacent images is statistically short), motion search is performed. It is preferable to narrow the range to improve the accuracy of the motion vector (to prevent a decrease in accuracy).

  The frame generation unit 906 generates and outputs a motion compensated image from the image supplied from the frame memory 901 and the motion vector II. The frame generation unit 906 outputs the image 3 read from the frame memory 901 as it is when the input image is used as the output image.

  An example of setting magnification that can be applied in the present embodiment will be given. As shown in FIG. 6, in a moving image with a frame rate of 60 Hz, there is a difference of 2.5 times in the amount of motion between the video material and the film material (24 Hz). It is preferable to set ka = 1 / 2.5 and kb = 2.5. In order to facilitate the processing of the reduction processing unit 903 and the enlargement processing unit 905, the reduction magnification can be set to ka = 1/2 and the enlargement magnification can be set to kb = 2.

  As described above, according to the present embodiment, the motion search range is adjusted so as to match the substantial frame rate (that is, the frame rate of the material (original image)) without increasing the processing amount of the motion detection processing in the motion vector detection unit 904. It is possible to detect an accurate motion vector.

  FIG. 10 is a block diagram showing a configuration example of a frame rate conversion apparatus according to another embodiment of the present invention. The frame rate conversion apparatus shown in FIG. 10 is the same as the frame rate conversion apparatus shown in FIG. 9, the reduction processing unit 903, an enlargement magnification for controlling the enlargement processing unit 905, an external CPU (Central Processing Unit) not shown in FIG. Is output. That is, the frame rate conversion apparatus illustrated in FIG. 10 includes a frame memory 1001, a reduction processing unit 1003, a motion vector detection unit 1004, an enlargement processing unit 1005, and a frame generation unit 1006. Since these components correspond to the frame memory 901, the reduction processing unit 903, the motion vector detection unit 904, the enlargement processing unit 905, and the frame generation unit 906 in FIG. 9, detailed descriptions thereof are omitted.

  The frame rate conversion apparatus according to the present embodiment controls a motion search range (substantially) by controlling an image reduction magnification and a motion vector enlargement magnification according to a frame rate (for example, 60 Hz, 50 Hz) of an input moving image. (Motion search range) is controlled. In particular, it detects the frame rate of the input video (for example, 60 Hz for the NTSC system and 50 Hz for the PAL system) and keeps it when the frame rate is a specified value (for example, 60 Hz), and when the frame rate is low (for example, 50 Hz). The actual motion search range is expanded. The frame rate of the input moving image can be detected from the input moving image signal.

  In this embodiment, when the input moving image is (50 Hz) regardless of whether it is 2-2 pulldown or 3-2 pulldown as in the embodiment using the film determination unit 902, the image reduction magnification is used. For example, ka is determined to be 50/60 and is transmitted to the reduction processing unit 1003. At the same time, the enlargement factor kb of the motion vector is determined to be, for example, 60/50 and is transmitted to the expansion processing unit 1005. The reduction processing unit 1003 reduces the image 1 and the image 2 according to the reduction magnification ka and outputs them as the image 6 and the image 7, respectively. The motion vector detection unit 1004 uses the specified motion search range from the images 6 and 7. The motion vector I is detected at (a prescribed size). The enlargement processing unit 1005 enlarges the motion vector I according to the enlargement magnification kb, and outputs the motion vector II to the frame generation unit 1006.

  As illustrated here, control is performed to expand the substantial motion search range only when the frame rate of the input video is low, and when it is high (when the moving distance of the object between adjacent images is statistically short). For this, it is preferable to make the motion search range narrower than that to prevent the motion vector accuracy from being lowered.

  As described above, according to the present embodiment, the motion search range is expanded to match the actual frame rate of the input moving image without increasing the processing amount of the motion detection processing in the motion vector detection unit 1004, and accurate motion can be achieved. It becomes possible to detect a vector.

  As another embodiment of the present invention, the frame rate conversion apparatus controls the motion reduction by controlling the image reduction magnification and the motion vector enlargement magnification according to the statistical value of the motion vector detection result in the motion vector detection means. The range (substantial motion search range) may be controlled. Here, the part where the statistical value is calculated is separately provided so as to obtain the output of the motion vector detecting unit 904 and pass the calculation result to the reduction processing unit 903 and the enlargement processing unit 905 in FIG. May be configured so that the external CPU in the description of FIG. 10 obtains and executes the output of the motion vector detection unit 1004. As described above, in the present embodiment, by using the detection result of the motion vector detection unit 904 or 1004, feedback control according to the actual motion amount is possible, and the result of the convergence to the actual motion amount to some extent. From this, the motion search range can be determined. Of course, in the present embodiment, the detection result may be obtained from the image before reduction instead of the detection result by the motion vector detection unit 904 or 1004.

  An example of controlling the reduction magnification and the enlargement magnification according to the statistical value will be given. For example, based on the statistics of the motion vector detection results, when there are many large motion amount distributions, the reduction ratio ka may be reduced and the enlargement magnification kb may be increased to substantially expand the motion search range. As another example, based on the statistics of the motion vector detection results, it is determined that an incorrect motion vector is detected beyond the motion search range when the motion vector distribution varies, and the reduction factor ka is reduced. The enlargement factor kb may be increased to substantially expand the motion search range.

  As described above, according to the present embodiment, an accurate motion vector can be obtained by expanding the motion search range so as to match the actual motion vector detection result without increasing the processing amount of the motion detection processing in the motion vector detection unit 904. It becomes possible to detect.

  As another embodiment, the above-described embodiments are combined to increase the motion detection process by controlling the motion detection range according to the input moving image material, the actual frame rate, and the motion vector detection result. The correct motion vector can be detected without any problem.

  In addition, the frame rate conversion device according to each of the embodiments described above according to the present invention is more effective when incorporated in a liquid crystal display device, but can also be applied to an image display device employing a panel of a display method other than liquid crystal, The same effect is produced. Furthermore, the frame rate conversion apparatus according to the present invention is mounted not only on an image display apparatus such as a television apparatus and a monitor but also on other AV equipment such as a recorder and STB (Set Top Box), the main body of a PC, etc. It can be assumed that the video output to the side is the video after the frame rate conversion. In that case, the image display apparatus on the output side may correspond to the frame rate.

  As described above, the illustrated frame rate conversion apparatus according to the present invention may be entirely configured by hardware, but a part thereof may be configured by software (including firmware). As an example for realizing such a configuration, when a computer such as a microcomputer in an AV device or a PC main body is used, generally, a hardware including a CPU, a memory, a bus, an interface, a peripheral device, and the like. And software that can be executed on these hardware (including a conversion program for executing the frame rate conversion process described above). The CPU controls the other hardware by executing this software. Specifically, the software stored in the ROM is expanded on the memory, and the expanded program is sequentially executed, so that the data on the memory (image data on the frame memory, etc.) or the interface is input. The functions of each unit are realized by processing, storing, and outputting data. The frame rate conversion process described above can be realized by incorporating and executing the conversion program for causing the CPU to execute the processing procedure as the software. The software may be stored in a recording medium and distributed, or distributed via a network such as the Internet or on a broadcast wave.

  Although the frame rate conversion apparatus of the present invention has been described above, the present invention can also take the form of a frame rate conversion method as described for the processing method. This method can be realized by each unit in the frame rate conversion apparatus, and can also be realized by executing the conversion program. The frame rate conversion method of the present invention will be briefly described.

  As described for the processing of the frame rate conversion apparatus, this frame rate conversion method detects a motion vector between images in an input moving image, and generates a new image using the detected motion vector. Convert the frame rate of moving images. In this frame rate conversion method, the image reduction means reduces the image in the moving image, and the motion vector detection means detects the motion vector between the images reduced in the image reduction step. A motion vector expanding step in which the motion vector expanding means expands the motion vector detected in the motion vector detecting step; and an image generating means in the moving image using the motion vector expanded in the motion vector expanding step. An image generation step for generating a new image from the image, and the control means controls the image reduction magnification in the image reduction step and the motion vector enlargement magnification in the motion vector enlargement step, thereby reducing the motion vector search range. Controlling.

  Here, the control means conveys information (film determination result, actual reduction magnification / enlargement magnification, etc.) for changing the reduction magnification and the enlargement magnification to the reduction processing unit and the enlargement processing unit in FIGS. Means. Other application examples of the frame rate conversion method of the present invention are the same as those described for the frame rate conversion apparatus, and the description thereof is omitted.

901, 1001 ... Frame memory, 902 ... Film determination unit, 903, 1003 ... Reduction processing unit, 904, 1004 ... Motion vector detection unit, 905, 1005 ... Enlargement processing unit, 906, 1006 ... Frame generation unit.

Claims (5)

  In a frame rate conversion device that converts a frame rate of the moving image by detecting a motion vector between images in the input moving image and generating a new image using the motion vector, the image in the moving image is An image reducing means for reducing, a motion vector detecting means for detecting a motion vector between images reduced by the image reducing means, a motion vector expanding means for expanding a motion vector detected by the motion vector detecting means, Image generation means for generating the new image from the image in the moving image using the motion vector enlarged by the motion vector enlargement means, and the image reduction magnification in the image reduction means and the motion vector enlargement means Control the search range of motion vectors by controlling the magnification of motion vectors. Frame rate conversion device according to symptoms.   The search includes a material detection unit that detects the material of the moving image, and controls a reduction magnification of the image and a magnification of the motion vector according to a detection result of the material of the moving image in the material detection unit. The frame rate conversion apparatus according to claim 1, wherein the range is controlled.   The frame rate conversion apparatus according to claim 1, wherein the search range is controlled by controlling an image reduction magnification and a motion vector enlargement magnification according to a frame rate of the moving image.   The search range is controlled by controlling a reduction magnification of an image and an enlargement magnification of a motion vector according to a statistical value of a motion vector detection result in the motion vector detection means. Frame rate conversion device. In a frame rate conversion method for converting a frame rate of the moving image by detecting a motion vector between images in the input moving image and generating a new image using the motion vector,
An image reduction means for reducing an image in the moving image;
A motion vector detecting step for detecting a motion vector between the images reduced in the image reduction step;
A motion vector expanding means for expanding a motion vector detected in the motion vector detecting step;
An image generating step for generating the new image from the image in the moving image using the motion vector expanded in the motion vector expanding step;
A control means for controlling a search range of a motion vector by controlling an image reduction magnification in the image reduction step and a motion vector enlargement magnification in the motion vector enlargement step;
A frame rate conversion method characterized by comprising:

Images