JP2007259477A - Image composition method, apparatus, and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To reduce a user operation upon generating a strobe composite image and shorten the time required for image generation process. <P>SOLUTION: A first frame is first selected from a moving image, before a plurality of frames of the moving image are superposed to generate a strobe composite image. Next, a plurality of image frames to be subjected to strobe compositing are determined. The plurality of image frames are determined, with reference to a frame of interest, corresponding to a position obtained by shifting the position on a time axis of the selected first frame by a predetermined value. Furthermore, a superposing method, when the plurality of image frames are superposed, is set. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to an image composition method, apparatus, and program for generating a strobe composite image by superimposing time series of subject images.

  Image processing technology that cuts out only the subject part from a moving image and overlays it with other images to generate and display a composite image is useful, for example, for generating a strobe composite image in which subject images taken at multiple times are superimposed It is. At that time, unless a user interface for easily superimposing a plurality of subject images is not provided, a strobe composite image cannot be easily generated, and time is spent on many operations.

  Conventionally, a technique of superimposing a plurality of image constituent elements called layers having a transparent part has been used to generate a strobe composite image in which a plurality of subject images are superimposed (see, for example, Non-Patent Document 1). First, only the subject image area is cut out from the moving image by some method in advance. For this, for example, as shown in FIGS. 28A to 28C, a chroma key method is used in which the motion of the object 1001 is photographed under the background 1002 of a uniform color on the entire surface, and cut out using the difference in color components. it can. Then, by using the layer function of the image processing software, the cut out subject image region is duplicated to form a plurality of layers, and the plurality of layers are overlapped to obtain a strobe composite image.

Specifically, for example, moving images composed of subject images taken at a plurality of times as shown in FIG. 29 are read, and a plurality of layers are created as shown in FIG. Next, as shown in FIG. 31, the layer 2 is shifted with respect to the layer 1 in terms of time, and the layer 3 is shifted with respect to the layer 1 in terms of time by a different amount. In this state, a plurality of frames can be superimposed on one frame. FIG. 32 shows a frame 501 at a certain time with three layers in FIG. 31, and FIG. 33 shows a state in which these layers in FIG. In this way, a strobe composite image can be generated by editing subject images taken at a plurality of times as separate layers and overlaying them.
Adobe Premier 6.0 Manual Chapter 7 pp281-294

  However, in order to create a strobe composite image with the conventional technology, a layer corresponding to each frame is created by the user's operation, and the overlay method is also designated by the vertical relationship at the stage of creating the layer, or layer creation Later, it was necessary to adjust the vertical relationship for each layer by a user operation. Further, in the conventional technique, it is necessary to capture an image that is a target of strobe composition once and save it in a file before performing strobe composition. In addition, for example, when shooting is not possible under a uniform background, such as in a sporting game, the chroma key method cannot be used to cut out the subject portion from the image, so the cutting work must be performed manually. Therefore, it takes a long time to create the strobe composite image.

  Strobe composite images are often used to explain athletes' forms in live TV broadcasts, and it is often necessary to shorten the time from when an image is obtained until the strobe composite image is created. It takes a lot of time to create a strobe composite image using the prior art, and these demands cannot be met. Therefore, there is a need for a method for generating a desired strobe composite image in a short time.

  The present invention has been made in view of such circumstances. An image composition method, an image composition apparatus, and an image that can reduce user operations during the generation of a strobe composite image and reduce the time required for image generation work. The purpose is to provide a synthesis program.

  The image composition method, apparatus, and program according to the present invention generate a strobe composite image by superimposing a plurality of frames of a moving image. A first frame is selected from the moving image. Next, a plurality of image frames to be combined with the strobe are determined. The plurality of image frames are based on a target frame corresponding to a position shifted by a predetermined number of positions on the time axis of the selected first frame. Further, a superposition method for superposing the plurality of image frames is set.

  According to the present invention, it is possible to reduce the burden of user operation necessary for generating a strobe composite image, and to significantly reduce the time required for the strobe composite image generation work.

  Embodiments of the present invention will be described below with reference to the drawings.

  (First Embodiment) FIG. 1 is a block diagram showing a schematic configuration of an image composition apparatus according to a first embodiment of the present invention. As shown in FIG. 1, the image composition apparatus according to this embodiment includes an input video and a display device 1 for displaying various screens for generating and editing a strobe composite image based on the input video, and a central processing unit (CPU). 2, an input device 3 including a keyboard and a pointing device (such as a mouse), and a main storage device 4. A program 5 for generating and editing a strobe image, input video data 6, and generated strobe image data 7 are recorded in an external recording device such as a hard disk device or a magneto-optical disk device.

  The image synthesizing apparatus of this embodiment can be realized as software that uses a general-purpose computer and operates on the computer. In this case, the computer program 5 responsible for processing for generating and editing the strobe image is read into the main storage device 4 and executed by the central processing unit 2. In the configuration of the present embodiment, an operating system (OS) that controls the hardware of the computer and provides a file system and a graphical user interface (GUI) is also introduced. In such a configuration, a computer program responsible for processing for generating and editing the strobe image is incorporated as application software that operates on the operating system.

  Before describing specific processing operations of the image composition apparatus according to the present embodiment, first, an outline procedure of strobe composition will be described.

  In strobe composition, an input image is a moving image, and an output image is a strobe-combined image (moving image or still image). In order to generate an output strobe composite image from an input moving image, a step of subject extraction that extracts only a subject portion to be strobe-combined from an input image, and which frame and how to overlap which frames are used for strobe composition Two steps of indicating the strobe composition target to be represented are required. These two steps are independent of each other, and either may be executed first.

  Here, the strobe-combined moving image Y refers to the input subject moving image as X, where the Y first frame is the X first frame image and the Y second frame is the Y first frame. An image obtained by superimposing the second frame of X on the frame, an image obtained by superimposing the third frame of X on the second frame of Y, an image obtained by superimposing the third frame of Y, and the (k + 1) (k) of Y Is a natural number) frame, an image obtained by superimposing the (k + 1) th frame of X on the kth frame of Y, and an appropriate time is given to each frame of Y to form a moving image. Such superposition is shown in FIG.

  In addition, although it is assumed that the frame is superimposed on the upper side here, frames that are instructed to be overlapped by the user are superimposed on the lower side. Similarly, even when only discrete frames are used, as described later, the images are superimposed above or below according to the time sequence of the input subject image. Further, the time series of the input subject image does not have to be the same as the time series when the subject is photographed, and may be, for example, the reverse order of the time series, so-called reverse reproduction.

  Further, the frames to be overlapped do not necessarily have to be continuous and may be discrete. In particular, since it is often desired to superimpose every N frames (N is an integer), a user interface (A, B, N) that targets N frames from the Ath frame to the Bth frame. Can be an integer).

  For example, 10 frames of moving images from frame number 0 to frame number 9 are subject to synthesis every two frames, such as frame number 1, frame number 3, frame number 5, frame number 7, and frame number 9. Thus, it is possible to generate a composite image from only odd frames of interlaced scan moving images.

  As another example, when the movement of the subject is not so violent, for example, as shown in FIG. In such a case, the strobe composite image obtained as shown in FIG. 4 becomes easy to understand by thinning out to a certain extent, for example, so as to be subject to synthesis every two frames.

  An interface for instructing the user to specify one frame for strobe image composition and further instructing superposition will be described. Based on this interface, a series of procedures as shown in FIG. 5 is executed.

  First, a frame selection input procedure (step S1) is performed, and it waits for one frame to be selected by the user. When a frame is selected, a frame shifting procedure is performed to calculate and determine a target frame commonly called an IN point or an OUT point (step S2). Thereafter, a superposition method setting procedure is performed on the target frame obtained by the frame shifting procedure (step S3). Finally, superposition is performed to obtain a strobe composite image (step S4).

  In the frame selection input procedure (step S1), for example, as shown in FIG. 6, an interface for displaying the slider 3201 and the frame 3202 being selected is presented to the user. In this interface, the user can select a desired frame by operating the slider 3201 with a mouse or a keyboard.

  A plurality of frames may be presented in the frame selection input procedure. For example, as shown in FIG. 7, the frame being selected (the frame of interest) and the frames before and after the frame are presented, and the frame can be selected not only by the selection by the slider but also by clicking the frame itself (1401 to 1403). If the interface is configured, the frame selection operation is more intuitive and easy for the user to use. Alternatively, it is also preferable to present a frame when a frame shifting procedure is performed on the currently selected frame. As a result, the user can immediately know which frame is designated by the current selection.

  In the superposition method setting procedure (step S3), for example, the target frame determined in step S2 is set to either the start point or the end point of strobe composition. When strobe compositing is performed by overlaying N frames (N is a natural number) from a specific start frame to a specific end frame in order (up or down) in order, strobe compositing the start frame And the end frame is the end point of strobe composition. It should be noted that the frames do not necessarily have to be continuous, and the starting point and the ending point may be determined similarly for an arbitrary number of discrete frames. Further, a plurality of frame shifting procedures and a superposition method setting procedure may be executed simultaneously based on the user's frame selection. For example, both the start point and end point of strobe composition may be determined simultaneously from one frame designation.

  Here, a method for specifying the start point of strobe composition will be described. FIG. 7 shows an example of a screen presented to the user for starting point designation. In the present embodiment, it is assumed that the stroboscopic composition is performed from N frames before the frame of interest to M frames, and N and M (integer values) are determined in advance. When the frame of interest 1402 is selected by the user (frame (number) 71 in this example), the frame before N frames (frame 65 in this example) is set as the start frame (start point), and the frame after M frames from the start frame It is considered that the overlaid strobe composition with the end frame (end point) as the end frame (end point in this example) is designated. When the “determine start point” button is pressed, composition is executed.

  Note that it is not always necessary to start the strobe composition when the “Start point determination” button is pressed. For example, when the “overlapping” button is pressed with the start frame to the end frame selected. It may be an interface that waits for a synthesis execution instruction. Further, the presented image may be a clipped subject or an image before cropping. In the above example, the end frame is defined, but M may not be determined, and only the start point may be selected.

  (Second Embodiment) Next, an interface for instructing a user to specify two frames for strobe image composition and further instructing superposition will be described. Based on this interface, a series of procedures as shown in FIG. 8 is executed. The first embodiment described above is preferable when a strobe composite image is quickly obtained by a minimum user operation. However, when the user wants to specify the start / end point (frame) of strobe composition strictly, the interface of this embodiment is used. Is useful. Further, in the present embodiment, with respect to the strobe composite image created once, an image can be generated again by changing the superposition method for some frames.

  First, a frame selection input procedure is performed, and in this embodiment, the user waits for two frames (start point and end point) to be selected (step S11). When a frame is selected, each frame between the two selected frames is used as a processing frame and superposition is performed (step S12). Here, in this embodiment, the superposition method is changed (step S13), and superposition is performed again to obtain a strobe composite image (step S14).

  For example, as shown in FIG. 9, the frame selection screen presentation procedure can be realized by presenting one slider 3301 and a frame 3302 selected by the slider 3301 and allowing the user to select two frames from the slider. A button 3303 and a button 3304 are buttons for setting the current attention point as a start point and an end point, respectively. FIG. 9 shows a state when the start point is set by pressing the button 3303. The user moves the slider 3301 as shown in FIG. 10, and sets the end point with the button 3304 as shown in FIG.

  As in the first embodiment described above, when the user selects a frame using the slider, not only the currently selected frame but also the frames before and after the selected frame are presented. If the frame can be selected by clicking on the frame, the frame selection operation becomes more intuitive and easy to use.

  The overlaying method in step S12 is arbitrary, for example, such that the frames are superposed in order from the oldest time, or conversely, the frames are superposed in order from the newest frame.

  Here, an example in which the designated range is overlapped will be described with respect to the series of procedures shown in the flowchart of FIG.

  An example of a screen presented to the user is shown in FIG. The purpose of the interface described here is to allow the user to select the first frame to be selected with the button 3004 with text and to select the last frame to be selected with the button 3005 with text. In FIG. 12, images of frames corresponding to buttons 3004 and 3005 are displayed on 3001 and 3003, respectively, so that the currently selected frame can be seen at a glance. The text 3006 displays the frame number of the current frame of interest, and the image 3002 displays the corresponding image. The user can move the point of interest by moving the slider 3013 to the left or right by mouse input or keyboard input. it can. The user can also instruct the superposition method using the exclusive push buttons 3010 and 3011 prior to the frame selection. These buttons are exclusive, and when any button is pressed, the other buttons return to the non-pressed state.

  FIG. 13 to FIG. 19 are diagrams showing the extracted buttons 3004, 3005, 3006 and slider 3013 in FIG. 12, and a series of selecting a target first frame and last frame with reference to these drawings. The operation procedure will be described. The display objects of the buttons 3004, 3005, and 3006 can display text characters representing the names of the buttons and numbers representing the frame positions (numbers) (hereinafter referred to as “buttons with text”). Note that the numbers on the buttons are updated in response to an operation on the slider 3013 and a confirmation instruction for the buttons.

  When the interface is first presented to the user, at this point, the start and end points are not specified. In this state, as shown in FIG. 13, the end point is displayed in gray so that the button cannot be pressed. Therefore, first, as shown in FIG. 13, the user moves the target point to the frame that the user wants to start. Next, as shown in FIG. 14, the user presses a button 3004 with text at the start point, and sets the current attention point as the start point.

  Next, as shown in FIG. 15, the user moves the attention point to the frame that is desired to be the end point. Since the target frame has not been determined at this time, the target frame is selected as a light color. When the button 3005 with text at the end point is pressed in this state, the end point is fixed, and the target frame accepts the superposition designation according to the push state of the exclusive push buttons 3010 and 3011 for designating the superposition method at that time. When the button with text 3005 at the end point is pressed again in the state of FIG. 16, the confirmation is released and the state returns to the state of FIG. Also, if the “execute composition” button 3014 is pressed in the state shown in FIG. 16, strobe composition is performed. In addition, when the frame of interest is selected again in the state of FIG. 16 and the start point button is pressed again as shown in FIG. 17, another strobe image of the new section is continued from the previous overlay designation as shown in FIG. Can be instructed to superimpose. The user selects an end point for this new section as shown in FIG. 19 and changes the pressing state of the exclusive push buttons 3010 and 3011 as necessary, and performs the same operation as the previous target frame selection, The overlay instruction can be given to the target frame.

  A button 3007 is a button for storing the current overlay setting in the storage device 4, and as shown in FIG. 20, the overlay setting for the frame at the relative position from the current point of interest is saved. If a point of interest is selected and a button 3012 is pressed, the stored overlay setting is read and applied. By recalling the saved settings in this way, the operation time can be shortened. Also, if you prepare a list of some of the settings that you have stored in advance, present that list as needed, and load other settings, you can use multiple settings separately. This is preferable because it can be done.

  Also, the current overlay setting stored in the storage device 4 may be recorded in a recording device such as an HDD, and the contents of the overlay setting may be read later and applied to another strobe image composition. . That is, the other strobe image may be synthesized by applying the overlay setting read out later to a frame group different from the currently processed one.

  In addition, a plurality of overlay settings with different contents are recorded in the recording device, and a strobe composite image for each overlay setting is generated at once, or a plurality of overlap settings are generated based on a certain overlap setting. A strobe composite image may be generated. It is preferable from the viewpoint of improving work efficiency that only the overlay setting operation is performed and the image generation processing is performed collectively at a later time.

  Next, an interface for generating and presenting a strobe composite image once with all the frames as an overlay, collecting objects indicated by the user, changing them to a bottom overlay, and generating a strobe composite image again will be described. To do. This corresponds to the procedure of step S13 and step S14 in FIG.

  As shown in FIG. 21, first, a composite image is displayed in an overlaid state (FIG. 21A), and the user designates the object 2101 of the left three subject images by left-clicking the mouse pointer on the screen. . The designated object 2101 may be colored so that it can be distinguished from other subjects (FIG. 21B). Then, when the user right-clicks, the three selected objects 2101 are switched to a lower layer, and a new composite image as shown in FIG. 21C is displayed. Note that the interface of FIG. 12 is presented in advance without presenting all frames as an upper layer in advance, and when the right click is performed, the upper layer is changed to the lower layer and the lower layer is changed to the upper layer. You can also In this way, it is possible to realize both the designation of a rough overlay and the adjustment of the overlay in units of frames in a short time.

  In this embodiment, the interface for selecting the target frame from the strobe composite image and separately specifying the change of the overlay method has been described. However, when the target frame is selected from the strobe composite image by the user, the superposition order is also described. By adding an interface that can change the setting, it is possible to further shorten the time required for creating and editing the strobe composite image. However, changing the strobe composite image at the moment of clicking requires a high processing capability of the image compositing device, and the operator needs skill to cope with the instantaneous screen change, so that it can be used properly according to the situation. It is good to make it.

  In addition, although it has been explained that a slider is used for frame selection, for example, several representative frames are presented, and one of these representative frames is selected by the user, and only frames close to the selected representative frame are presented. If the user is asked to give an instruction for superposition based on these frames, the slider operation as described above can be omitted.

  As described above, by providing an interface that greatly simplifies user operations necessary for strobe composition, the strobe composition execution procedure can be clarified and the operation burden on the user can be greatly reduced. Therefore, the time required for creating the strobe composite image can be shortened, and the usability of the image composition system can be improved.

  (Third Embodiment) Next, a third embodiment provided with an interface for presenting a preview during a manual cutting operation will be described with reference to FIGS. With regard to the input video as shown in FIG. 22, the situation in which the user is cutting out the subject image, specifically, the cutting out of the first frame and the second frame is completed, and the cutting out of the third frame is performed. Think about the situation.

  As shown in FIG. 23, when the user cuts out the subject image 1702 so as not to include the subject image 1701, if the part 1801 of the image is cut out incorrectly as shown in FIG. The influence appears as a part 1901 of the strobe composite image as shown in FIG.

  Therefore, as shown in FIG. 26, if the image 1803 after the stroboscopic composition based on the current clipped image 1802 is sequentially synthesized and displayed as a preview as well as the input image and the clipped image, the user is cut out inaccurately. A portion that affects the strobe composite image can be immediately confirmed, and correction for obtaining a desired strobe composite image is facilitated. As shown in FIG. 27, the corrected strobe composite image can be immediately confirmed.

  Note that the preview is to present a processing result based on the current instruction content before the processing, and for example, a completed image (a still image or a moving image) may be displayed according to an input such as a user's overlay frame selection. It means to present to the user while updating sequentially.

  Note that it is not necessary to present all of the input image, the cutout image, and the strobe composite image. For example, only the input image and the strobe composite image may be presented. In addition, the preview may be presented not only when the user cuts out completely, but also when the user corrects the result of the automatic or semi-automatic cutting out by software.

  In addition, when a subject is cut out from an image area that does not change with the passage of time, such as the background, even if inaccurate clipping is performed, there is no effect on the strobe composite image. According to the preview, the user can confirm this. Therefore, it is possible to reduce the time for generating the strobe composite image without cutting out.

  Further, the user can know the final image during editing, and after the strobe composite image is generated, it is not necessary to check whether the strobe composite image is as expected by the user. Thus, sequentially presenting the strobe composite image as a preview greatly contributes to reducing the time required to generate the strobe composite image and reducing the burden on the user more than presenting the cutout image as a preview.

  (Fourth Embodiment) FIG. 34 is a flow chart showing a series of processing for strobe image composition according to a fourth embodiment of the present invention. In this embodiment, when an instruction from the user (eg, pressing a button, clicking a mouse, etc.) is received during moving image shooting, a desired composite image is generated in consideration of the timing. As an input device, for example, a camera for shooting a moving image is connected to a strobe composition system. Hereinafter, a system that implements the fourth embodiment is referred to as a one-instruction strobe composition system. A queue (first-in first-out buffer) for storing moving images is prepared in advance. A multi-stage delay circuit can be used instead of the queue.

  The one-instruction strobe composition system repeats steps S3401 to S3402 shown in FIG. 34 unless there is an instruction by user input. That is, the image input step (S3401) for fetching the next image frame from the input device and the image holding step (S3402) for discarding the first frame of the queue and shifting it by one frame and adding the frame fetched to the final frame are repeated. .

  Here, when an instruction by a user input (S3403: user input step) is detected, the one-indicating strobe composition system generates a timer that issues at least one notification (timeout) when a predetermined time has elapsed from the time when the user input occurred. Set (S3404) and continue the processing of S3401 to S3402. At the time of notification from the timer (S3405), strobe composition is performed using images stored in the queue (S3406: image composition step), and a composite image is output (S3407: image output step). According to the present embodiment, since the user only needs to give a single instruction, the burden on the user is greatly reduced, and the time required for the strobe composite image generation work can be greatly shortened.

  Fifth Embodiment FIG. 35 is a flow chart showing a series of strobe image composition processing according to the fifth embodiment. In the present embodiment, a frame (referred to as a “feature image frame”) that is important for strobe composition is automatically detected during moving image capturing, and a desired composite image is generated in consideration of the timing. Such a system of the fifth embodiment will be referred to as a “fully automatic strobe composition system”.

  As in the fourth embodiment described above, a queue (first-in first-out buffer) for storing moving images is prepared. The fully automatic strobe composition system according to the present embodiment repeats the following steps S3501 to S3503. That is, an image input step (S3501) for fetching the next image frame from the input device, an image holding step (S3502) for discarding the first frame of the queue and shifting it by one frame, and adding the fetched frame to the final frame. This is a feature image frame detection step (S3503) for checking whether an image frame exists.

  Here, if a feature image frame exists, the fully automatic strobe composition system sets a timer that issues a notification (timeout) only once a predetermined time has elapsed from the detection time of the feature frame (S3504), and the processing of S3501 to S3503 Continue. When a notification is received from the timer (S3505), strobe composition is performed using the images stored in the queue (S3506: image composition step), and a composite image is output (S3507: image output step).

  According to the present embodiment, it is possible to completely omit the instruction from the user, and the time required for the strobe composite image generation work can be further shortened as compared with the fourth embodiment described above. In the fourth and fifth embodiments in which a user instruction or a feature image frame is detected, an example using a timer has been shown. However, the time is recorded at regular time intervals to determine the passage of a predetermined time. May be. In the fully automatic strobe composition system, various methods can be considered as to how to define a feature image frame and how to detect it. A description will be given taking a strobe composite image in a golf tee shot as an example. As the strobe image composition matching condition, for example, the following can be considered.

(1) A sound generated during a tee shot is detected, and an image frame in which the sound is detected is set as a feature image frame.
(2) The shape (template) of the subject in the feature image frame is prepared in advance, the subject region is extracted for each image frame, and it is determined that the subject shape is sufficiently similar to the template. Let it be an image frame.
(3) A feature image frame and templates of subjects for several frames before and after it are prepared in advance, subject areas are extracted for each image frame, and it can be determined that the shapes of subjects for several consecutive frames are sufficiently similar to the template. Sometimes it is a feature image frame.
(4) When the area of the subject area is calculated for each image frame, the difference (peak / peak value) between the maximum and minimum areas of several consecutive frames is obtained, and the peak / peak value exceeds a separately determined value In addition, a frame that gives a maximum value (or a minimum value) of the area is defined as a feature image frame.

  In the extraction of the subject area, for example, a chroma key method in which a monochrome background is prepared in advance and a color area different from the background color is used as the subject area can be used. In calculating the area of the subject region, for example, the subject region may be extracted using the chroma key method, the number of pixels in the region may be calculated, and this may be used as the area. Further, as a reference for determining the similarity to the template image, for example, the area St of the template subject area, the area Se of the extracted subject area, and the area Sc of the area where the subject area of the template overlaps the extracted subject area are used. What is necessary is just to calculate each, and just to judge that it is close enough when the smaller one of the two values (Sc / St) and (Sc / Se) is equal to or greater than the threshold value.

  From these methods, if a method suitable for a target subject for strobe synthesis is selected in advance and a feature image frame is detected by the selected method, the detection accuracy of the feature image frame is improved. Therefore, the usefulness of the fully automatic strobe synthesis system is also improved. As a result, it is possible to reduce the time required for the strobe composite image generation work.

(Preset present)
In the one-instruction strobe composition system or the fully automatic strobe composition system, the quality of the strobe composition generated depends on the type of input image and the strobe composition algorithm. For example, when extracting the subject area of each target frame by the chroma key method, it is necessary to control a threshold value for setting which range is considered as the background color in the color space. If parameters such as threshold values in the chroma key method can be set by the user, high quality can be obtained. However, setting the parameters each time during strobe composition increases the burden on the user and increases the work time required for strobe composition. Therefore, it is preferable to prepare several types of parameters in advance so that the user can select the parameters to be used before the strobe composition. Thereby, the quality of the strobe composite image can be improved and the work efficiency can be improved.

(Present strobe frame or playback video)
Also, depending on the input image, it may be difficult to determine strobe composition parameters in advance. In that case, the user can obtain a desired output image by limiting the number of candidate parameters to several types, performing strobe composition on all of them, presenting the results, and allowing the user to select them. For example, in a user interface using a mouse and a screen, as shown in FIG. 36, strobe synthesis is performed with three parameters, and a desired strobe is operated by mouse operation from the final frame of strobe synthesis presented in 3601, 3602, 3603. Let the user select a composite image. Note that instead of presenting the final frame, a strobe-combined moving image may be reproduced. According to this method, even when sufficient quality cannot be achieved without parameter setting by the user, the work can be minimized, and as a result, the time required for strobe composition can be shortened.

  (Sixth Embodiment) FIG. 39 is a flowchart of strobe image composition according to the sixth embodiment of the present invention. In the present embodiment, automatic cut-out parameters used for strobe image synthesis are determined in advance. In this embodiment, a sports game is taken as an example, and a case where a sports player is photographed during the game and a strobe image is synthesized will be described.

  First, before a game, a moving image of a player (subject) is photographed for correct data and taken into a personal computer (S4101). Next, the moving image is displayed stationary on the display of the personal computer frame by frame, and the player's area is accurately input for each frame using a mouse or the like (S4102). Then, a strobe composite image is generated by combining the images obtained by cutting out the player's area while superimposing one frame at a time (S4103). This is a strobe composite image with extremely high synthesis quality, and this is called a correct strobe composite image.

  Next, the player area is automatically cut out from the same correct data image. An example of the cutting method is shown in FIG. A difference in luminance between the target frame 4301 to be cut out and another reference frame 4302 is obtained. If this inter-frame difference is larger than the threshold value, an object region is determined, and if not, an alpha mask 4303 is generated. The alpha mask is a bitmap composed of binary values of an object area and a background area.

  Next, a plurality of alpha masks 4305 based on differences from different reference frames 4304 are generated, and an AND of the object areas is obtained to obtain an accurate object area alpha mask 4306. Finally, the method described in the literature "Contour Fitting by Self-Similar Mapping" (Ida et al., Proc. Of the 5th Image Sensing Symposium, C-15, pp.115-120, June 1999.) If contour fitting is used, a more accurate alpha mask 4307 can be obtained. The contour fitting is a contour extraction process for obtaining a contour with fewer errors from a contour with large errors.

  In the cutting method as described above according to the present embodiment, the threshold value for determining whether the object area or the background area, the frame interval between the target frame and the reference frame, the strength of contour fitting, and the like are the cutting parameters.

  The extraction parameter is temporarily set to a certain value, and the player area is extracted (S4104). Then, a temporary strobe composite image is generated based on the player region cutout result (S4105). The temporary strobe composite image is generally inferior in composite quality than the correct strobe composite image. In order to measure the degree, an error of the temporary strobe composite image with respect to the correct strobe composite image is calculated (S4106). As parameters, some combinations may be determined in advance, and S4104 to S4106 may be repeated while switching within the range of the combinations. That is, if all the predetermined parameters have been executed, the process proceeds to the next step, and if not, the process returns to S4104 (S4107).

  After the above-described steps S4104 to S4106 are repeated for all parameters, the parameter with the smallest error of the temporary strobe composite image with respect to the correct strobe composite image is formally determined as a parameter to be applied thereafter (S4108). The parameters determined here are referred to as execution parameters in this embodiment. Then, from the image other than the correct data image photographed during the match, it is cut out using the execution parameter (S4109), and a strobe composite image is generated (S4110).

  The optimum value of the cut-out parameter depends on changes in the environment such as the color of the player's uniform, the background color, and the characteristics of noise generated during camera shooting. However, during the same game, these environments are constant, and good composition is possible without switching parameter values. According to this embodiment, it is possible to determine a parameter value that provides a good composite image even during a game. As a result, automatic compositing that does not require manual work is possible, and a strobe composite image can be generated in a short time.

  Note that a feature of this embodiment is that a composite image is used as an error evaluation object, instead of the shape of the clipped region. This is because, as described in the third embodiment, in the case of strobe image composition, even if the shape of the region is incorrect, the composite image may not be affected. Even if it is actually detected as a player's area where it is the background part, the same background is overwritten on the original background part without any change, and the same composite result as if it was not detected as a player area become. Further, as an error, for example, a difference between corresponding frames can be obtained, and the sum of absolute values or the sum of squares thereof can be used.

  (Seventh Embodiment) FIG. 40 is a block diagram showing a schematic configuration of a strobe image composition apparatus according to a seventh embodiment of the present invention. The seventh embodiment relates to a specific apparatus that executes a series of strobe image composition processes described in the sixth embodiment.

  For example, in the case of a personal computer, the photographed image 4201 is stored in a storage unit 4202 such as a semiconductor storage element or a magnetic recording device. The image 4203 is sent to a setting unit 4204 that sets a correct subject area as an image for correct data. Here, for example, the image 4203 is displayed on the display of the personal computer, and the user inputs the subject area accurately using the mouse or the like. The correct subject image 4205 cut out in that area is sent to the synthesis unit 4206. The composition unit 4206 overwrites the subject image 4205 in time order, and sends the correct strobe composite image 4207 to the holding unit 4208.

  On the other hand, the image 4203 is also sent to the subject region detection circuit 4212. The detection unit 4212 detects a subject area based on the cut parameter value 4211 sent from the parameter setting unit 4210, and sends a subject image 4213 cut out in that area to the combining unit 4214. The synthesizing unit 4214 overwrites the subject image 4213 in time order, and sends the strobe synthesized image 4215 to the error detecting unit 4216. Further, the correct strobe composite image 4207 is sent from the storage unit 4208 to the error detection unit 4216. The error detection unit 4216 calculates an error 4217 between the correct strobe composite image 4207 and the strobe composite image 4215. This error 4217 is sent to the parameter setting unit 4210.

  The parameter setting unit 4210 sequentially switches the parameter value 4211, examines an error 4217 obtained as a result, and obtains a parameter value that minimizes the error 4217. When another image 4201 is input, the parameter value 4211 is sent to the detection unit 4212, and the detection unit 4212 detects the subject area from the image 4201 read from the storage unit 4202. The subject image 4213 is combined with a strobe image by the combining unit 4214. This strobe composite image 4215 is output to the outside.

  With such a configuration, an automatic process is performed from an image input after the second time, and a strobe composite image can be generated in a short time.

  (Eighth Embodiment) The eighth embodiment of the present invention is configured such that the user can select an image frame to be used for synthesizing a strobe image while watching a slow playback video of a subject moving image. The eighth embodiment is executed from step S3701 according to a series of procedures shown in FIG. Before describing the contents executed in each step, the term “overlapping method switching frame” is used as follows. In other words, when generating a strobe composite image by superimposing multiple frames of a moving image, an image frame that switches the superposition method of whether to superimpose over or over superimpose is called a “superposition method switching frame”. To. The contents executed in each step are as follows.

  Step S3701: One image frame is captured and displayed. Thereafter, the process proceeds to step S3702.

  Step S3702: An image frame selection instruction from the user is accepted. If there is no instruction within the predetermined time, the process returns to step S3701. If there is an instruction, the process proceeds to step S3703.

  Step S3703: The image frame displayed in Step S3701 immediately before the user instruction received in Step S3702 is determined as a superposition method switching frame, and a position shifted by a predetermined number in the time direction from the position of the superposition method switching frame. Is determined as a start point, a position shifted by a separately determined number is determined as an end point, and a target image frame whose image frames at a predetermined number of intervals from the start point to the end point are subjected to strobe composition Determine as. Thereafter, the process proceeds to step S3704.

  Step S3704: If there are image frames that have not yet been captured among the image frames to be subjected to strobe composition determined in step S3703, they are captured. Thereafter, the process proceeds to step S3705.

  Step S3705: The image frames to be subjected to the strobe composition determined in step S3703 are superimposed to perform strobe composition. Then, this series of procedures is terminated.

  Conventionally, when a user tries to select a desired image frame while watching a playback video at a frame rate specific to the moving image of the subject, there are many cases where the user cannot select it correctly. become able to. As a result, the time until the strobe composite image generation can be shortened.

  The present embodiment can be modified as follows.

(1) If the user gives an instruction to accept in step S 3702 by mistake, the error cannot be corrected unless the series of procedures is completed and the procedure is started again. In order to be able to correct such an error, the image displayed when the user notices the error can be returned to the image displayed in the past, and the operation that has been erroneously performed can be performed retroactively. It is preferable to do so.

(2) In step S3701, an image is displayed. Here, if the user can determine the image display time, errors by the user in the instruction accepted in step S3702 can be reduced. It is also possible to change the display time during image display.

(3) One image frame is sequentially captured in step S3701, and a necessary number of image frames are also captured in step S3804. Instead of this, for example, image frames may be fetched at once, or a plurality of image frames may be fetched in step 3701.

(4) In accordance with the user instruction received in step S3702, the superposition method switching frame, start point, and end point are determined in step S3703. At that time, an image frame corresponding to the user instruction is determined as a superposition method switching frame. For example, an image frame corresponding to a user instruction may be determined as a start point or an end point.

(5) When it is not necessary to switch between the upper overlay synthesis and the lower overlay synthesis, it is not necessary to determine the overlay method switching frame, so the start point and the end point are set based on the image frame corresponding to the user instruction. It may be determined automatically. In this way, the number of user operations can be reduced, and the time until the strobe composite image is generated can be shortened.

(6) The superimposition method switching frame, the start point, and the end point are determined in accordance with one user instruction received in step S3702. For example, in order to obtain a strobe composite image desired by the user, One image frame may be determined by a user instruction instead of a plurality. This is effective when a strobe composite image with sufficient quality cannot be generated with a single user instruction, and this embodiment can be applied to more images.

(7) In the situation where only the start point has been determined and the end point has not been determined, when the execution of step S3705 is started immediately after the start point is determined, the time from the end point determination to the end of the strobe composition can be shortened.

  (Ninth Embodiment) In the ninth embodiment of the present invention, the user can select an image frame while viewing the reproduced video of the subject strobe synthesized moving image. The ninth embodiment is executed according to a series of procedures shown in FIG. This series of procedures is executed from step S3801. The contents executed in each step are as follows.

  Step S3801: One image frame is captured and displayed. Thereafter, the process proceeds to step S3802.

  Step S3802: An instruction to start strobe composition by the user is accepted. If there is no instruction to start within the predetermined time, the process returns to step S3801. Alternatively, if there is an instruction not to start, the process returns to step S3801. If there is an instruction to start, the process advances to step S3803.

  Step S3803: One image frame is captured. Thereafter, the process proceeds to step S3804.

  Step S3804: Accepting a user instruction whether to change the superposition method of whether to superimpose images or to superimpose them when superimposing image frames and combining them with flash. If there is a change instruction, the process advances to step S3805. If there is no change instruction within the predetermined time, the process advances to step S3806. Alternatively, if there is an instruction not to change, the process proceeds to step S3806.

  Step S3805: When superimposing image frames and combining them with a stroboscope, the superposition method of whether to superimpose or superimpose is changed. Thereafter, the process proceeds to step S3806.

  Step S3806: The image frames are superposed according to the superposition method to stroboscopically compose. Thereafter, the process proceeds to step S3807.

  Step S3807: The strobe composite image synthesized in step S3806 is displayed. Thereafter, the process proceeds to step S3808.

  Step S3808: A user instruction whether to end the strobe composition is accepted. If there is no instruction to end within the predetermined time, the process returns to step S3803. Alternatively, if there is an instruction not to end, the process returns to step S3803. When there is an instruction to end, the series of procedures is ended.

  According to this embodiment, strobe composition, strobe composition moving image reproduction, and image frame selection can be performed at the same time, so that the time required for completion of user-desired strobe composition can be shortened.

  The ninth embodiment can be improved as follows.

(1) If the user mistakenly gives an instruction to accept in step S3802, step S3804, and step S3808, the error cannot be corrected unless the series of procedures is completed and the procedure is performed again. In order to be able to correct such an error, the image displayed when the user notices the error can be returned to the image displayed in the past, and the operation that has been erroneously performed can be performed retroactively. Like that.

(2) An image is displayed in steps S3801 and S3807. If the user can determine the time for displaying these images, errors by the user in the instructions accepted in steps S3802, S3804, and S3808 can be reduced. It is also possible to change the display time during image display.

(3) Although one image frame is sequentially fetched in each of steps S3801 and S3803, for example, image frames may be fetched all at once, or a plurality of image frames may be fetched in step 3801 and step S3803. Also good.

(4) When a user instruction is received in step S3804, the superposition method at the time of strobe composition is changed. A position shifted in the time direction from the position of the image frame corresponding to the user instruction may be determined as the end point. In that case, the end points are superimposed and combined, and when the strobe composite image is displayed, the series of procedures is ended. As a result, the number of instructions given by the user can be reduced by one, and the time until the end of strobe composition can be shortened.

(5) The position shifted in the time direction may be determined as the start point instead of being determined as the end point. In this way, if an image frame at a position shifted in the time direction from the position of the image frame corresponding to a certain user input is determined as a frame for changing the start point, end point, and superposition method, it is accepted in step S3802. Even if the user instruction is incorrect, there is no need to redo or correct the operation, and the time required for strobe composition can be reduced.

(6) In the ninth embodiment, a strobe composite image is displayed simultaneously with strobe composition. The strobe composite image may be stored in a storage device so that the strobe composite image can be reproduced later.

  (Tenth Embodiment) FIG. 42 is a flowchart showing the flow of processing of a strobe image composition method according to a tenth embodiment of the present invention. In FIG. 42, S4601 is an image input step for sequentially inputting time-series image frames, S4602 is an image holding step for holding the latest N frames (N is a natural number) input by the image input step, and S4603 is a strobe. A reference image frame detection step for detecting a reference image frame serving as a reference for a plurality of image frames to be combined with a stroboscopic image according to a stroboscopic image synthesis matching condition for determining whether or not the image frame is suitable for image synthesis. S4604 is a reference image. An image compositing step for performing a strobe image compositing process on a plurality of image frames stored in the image holding step when a separately determined time has elapsed since the detection of the frame, S4605 is a strobe image compositing by the image compositing step Processed composite image frame An image output step of outputting a beam.

  The image input step S4601 corresponds to a step of shooting and inputting a subject to be stroboscopically combined, for example, camera shooting at a sports broadcast. Here, time-series image frames are sequentially input. The image holding step S4602 is a step of holding the latest N frames of image frames input in the image input step S4601, and for example, for example, for example, a sports broadcast video while constantly updating the holding contents of a limited capacity storage medium. Hold the latest 2 seconds (60 frames for 30 [fps]).

  In the reference image frame detection step S4603, whether or not the currently held image sequence is an image frame suitable for strobe image synthesis in a state where the latest N frame image frames are held in the image holding step S4602. A reference image frame serving as a reference for a plurality of image frames to be combined with a stroboscopic image is detected according to the determined stroboscopic image composition matching condition. For example, when a baseball broadcast video is input in the image input step S4601, an image sequence to be subjected to the strobe composition processing is a pitcher throwing scene, a batter batting scene, or a multiple image frame of a ball. Here, the reference image frame corresponds to the frame at the moment when the pitcher releases the ball and the frame at the moment when the batter hits the ball, and a desired image can be created by combining the frames before and after the strobe image. The strobe image composition matching condition includes, for example, a condition using image information and a condition using information different from the image. In conditions using image information, information only in the image is used. On the other hand, under conditions using information different from the image, information different from the image is obtained by another input step synchronized with the image input step. For this, a microphone, an infrared sensor, a weight scale, or a physical switch is used. Under these two conditions, for example, the reference image frame can be detected as follows.

(Conditions using image information)
A frame in which an image frame similar to a previously registered image is detected is detected as a reference image frame. The pre-registered image is, for example, an image frame at the moment of impact.

A frame in which a change in pixel value of a partial area set in advance in the frame is detected is detected as a reference image frame. For example, a frame in which a position where a golf ball is placed is set as a partial area and a white pixel changes to a green pixel, or a baseball bat passes through a certain area.

-When a mask image similar to a preset mask image (binary image used for separating and extracting a subject area and a background area) is created, the original image frame of this mask image is detected as a reference image frame To do.

In the mask image created from the image frame, a frame having the smallest subject area is detected as a reference image frame. Baseball batting and golfer swings tend to minimize the area of the subject area at the moment of impact.

(Conditions using information different from images)
When a preset sound is detected, a frame corresponding to that time is detected as a reference image frame. Examples of the detection sound include baseball batting sound and golf impact sound.

-When the infrared sensor senses an object to be detected, the frame corresponding to that point is detected as a reference image frame. For example, when an infrared sensor detects that a baseball bat, pitcher's arm, or golf club has passed a predetermined point, it is detected that there is no longer a ball on the golf tee or artificial grass mat. This is an image frame corresponding to the time point.

When a change in weight is detected, a frame corresponding to that point is detected as a reference image frame. The frame at the time when the weight of the golf tee or the artificial turf mat changes may be suitable as the reference image frame.

When a physical switch is operated, a frame corresponding to that point is detected as a reference image frame. For example, when a pitcher steps on a pitching plate, when a pressure is applied to a golf tee or an artificial grass mat.

  In the image composition step S4604, strobe image composition is performed on the plurality of image frames stored in the image holding step S4602 based on the reference frame detected in the reference image frame detection step S4603. At this time, the strobe composition may be performed on the plurality of image frames stored in the image holding step S4602 immediately after the reference image frame detection step S4603 detects the reference image. Alternatively, strobe composition may be performed on a plurality of image frames stored in the image holding step S4602 after a predetermined time has elapsed since the reference image frame was detected. The former is a case where the reference image frame is detected as the last frame of a plurality of frames to be subjected to strobe composition, and the latter is a frame in the middle of the plurality of frames to be subjected to strobe composition (for example, , The moment of impact of baseball batting or golf swing). The number of frames to be combined with the flash may be all the images held in the image holding step S4602, or a predetermined number of frames, for example, (2M + 1) frames of M frames before and after the reference image frame as a center. You may synthesize.

  In the image output step S4605, the composite image frame that has been subjected to the strobe composition processing in the image composition step S4604 is output on a display or as a video file.

  As described above, detecting a reference image frame that serves as a reference for a plurality of frames to be combined with a stroboscopic image based on a stroboscopic image combining condition that determines whether or not the image frame is suitable for the stroboscopic image composition. This means that the section designation of the plurality of image frames can be automatically performed. As a result, the time from shooting to strobe composite video output can be shortened, and a system that performs strobe composite output without operator processing can be realized.

  (Eleventh Embodiment) FIG. 43 is a flowchart of a strobe image composition method according to an eleventh embodiment of the present invention. In FIG. 43, 4701 is input for a strobe composite image and is an input video composed of a plurality of image frames representing the subject, 4702 is a feature point tracking step for obtaining the movement locus of the feature point of the subject, and 4703 is the obtained locus. A motion pattern determination step for analyzing the motion of the subject based on the pattern, 4704 is a motion pattern determination result determined by the motion pattern determination step.

  In a feature point tracking step 4702, feature points of the subject photographed in the input video 4701 are tracked. This can be done manually by a user using a pointing device, or feature points can be automatically tracked using color information of an image. For tracking, for example, a histogram distribution method can be used in which a color distribution histogram is created for a rectangular area around a feature point, and an area having the closest color distribution histogram is searched for each frame. When the input video is a video that has already been stroboscopically synthesized, a plurality of feature points are synthesized into a single image frame, making it easier to track the feature points. As an example of the tracked feature points, the results of tracking the club head and the golfer's head during a golf swing are shown in FIGS. 44 (a) and 44 (b). FIG. 44B shows an enlarged view of the vicinity of the head of the subject shown in FIG. 44A, and shows the trajectory of the golfer's head.

  45 (a) and 45 (b) are diagrams showing how the motion pattern determination step is executed. In the motion pattern determination step 4703, motion analysis of the subject is performed based on the pattern of the feature point tracking result that is the trajectory of the feature point obtained in the feature point tracking step 4702. FIG. 45A shows the case of analyzing the movement pattern of the club head, and FIG. 45B shows the case of analyzing the movement pattern of the golfer's head. 45A and 45B, reference numerals 4901 and 4904 denote inputs to the motion pattern determination step 4703, which correspond to the feature point tracking results output from the feature point tracking step 4702, respectively. Reference numerals 4902 and 4905 denote motion pattern determination units that determine a motion pattern by comparing a pattern registered in advance with a pattern of a feature point tracking result. Reference numerals 4903 and 4906 denote output examples of the motion analysis results determined by the motion pattern determination units 4902 and 4905.

  The golf club head trajectory pattern is such that when the player is viewed from the side, the follow-through swing trajectory is generally larger than that during the downswing. It is said that it is a good swing using the weight of the head. Such an exercise pattern of the club head representing a good swing, an exercise pattern of a club head representing a bad swing, an exercise pattern of a club head representing a swing that is likely to be experienced by beginners, and the like are registered in the exercise pattern determination unit 4902 in advance. Then, for the exercise pattern input 4901, the most similar pattern is selected from the registered exercise patterns, and the exercise analysis result 4903 is output. In the example of FIG. 45A, a motion analysis result 4903 “advanced” is obtained for the input motion pattern 4901.

  In addition, when looking at the player from the side, the trajectory pattern of the golfer's head is considered to be a stable swing with a stable spine axis because the upper body is not swayed when there is less blur in the horizontal and vertical directions. The

  In the swing of an advanced player, the weight moves to the left foot toward the impact of hitting the ball from the state where the weight is moved to the right foot by the takeback (the state where the locus is moved to the left end in FIG. 45B). (The trajectory moves to the right as 4908 in the figure), and immediately before the impact, the club head and the arm are momentarily pulled to accelerate the club head (at this time, the trajectory at the top of the head again in the figure) Move to the left), and hit the impact with Behind Ball 4909. After the impact, the body weight moves to the left foot by natural weight shift (4910). This is the movement pattern of the top of the head in the swing of the advanced player.

  As in the case of the club head described above, an exercise pattern determination unit is used for an exercise pattern of the top of a senior, a movement pattern of the top of the head indicating a bad swing, an exercise pattern of the top of the head representing a swing that is likely to be a beginner. Register in advance in 4902. Then, for the motion pattern input 4904, the most similar pattern is selected from the registered motion patterns, and a motion analysis result 4906 is output. In the example of FIG. 45B, a motion analysis result 4903 “advanced” is obtained for the input motion pattern 4904.

  According to the present embodiment described above, in an image composition method for generating a strobe composite image by superimposing a plurality of frames of a moving image, a movement locus of a feature point of a subject is obtained, and the obtained locus pattern is used as a basis. By performing the motion analysis of the subject, it is possible to output not only the strobe composite image visually but also the motion analysis result that leads to the improvement of the motion technique and the like.

  (Twelfth Embodiment) FIG. 46 is a flowchart of a strobe image composition method according to a twelfth embodiment of the present invention. In FIG. 46, S5001 is an object region extraction step for extracting an object region in real time from an image frame being shot, S5002 is a reference frame specification step for specifying an image frame serving as a reference for strobe composition, and S5003 is a specified reference frame. This is an image composition process for performing strobe composition using a cut object region.

  First, in the object region cutout step S5001, the object region is always cut out in real time regardless of whether or not the input image is a frame for which the strobe composition processing is performed. In the reference frame specifying step S5002, for example, a user's manual specification or an image frame serving as a reference for strobe composition is specified by an automatic method as described in the tenth embodiment. In the image composition step S5003, strobe composition is performed using the designated reference frame and the clipped object region. In the strobe image compositing method described so far in the above embodiment, after specifying an image frame as a reference for strobe compositing, the object region is cut out from a plurality of frames to be strobe composited to perform strobe compositing. . This has a problem that the object region cut-out processing time is also added from the determination of the output video range to the actual video output. However, as described in the present embodiment, this problem can be solved when the object region cutout process is always performed in real time.

  In a stroboscopic image synthesis method for generating a stroboscopic image by superimposing a plurality of frames of moving images, an object region is cut out in real time from an image frame being shot, and an image frame serving as a reference for stroboscopic synthesis is awaited. When the reference image frame is designated, strobe composition is performed using the designated reference frame image and the clipped object region. Therefore, the time from shooting to strobe composite video output can be greatly shortened.

  Note that the present invention is not limited to the above-described embodiment as it is, and can be embodied by modifying the components without departing from the scope of the invention in the implementation stage. In addition, various inventions can be formed by appropriately combining a plurality of components disclosed in the embodiment. For example, some components may be deleted from all the components shown in the embodiment. Furthermore, constituent elements over different embodiments may be appropriately combined.

1 is a block diagram showing a schematic configuration of an image composition device according to a first embodiment of the present invention. Diagram showing how frames are superimposed Figure showing a strobe composite image synthesized from a slow-moving object image Diagram showing strobe composite image generated by thinning out some frames Flow chart showing a series of procedures for changing the superposition method by designating one frame The figure which shows an example of the attention point selection interface by a slider The figure which shows an example of the screen shown to a user for the designation | designated of a synthetic | combination start point A flow chart showing a series of procedures for specifying a target range and changing a superposition method according to the second embodiment. The figure which shows an example of the selection interface of two frames of a start point and an end point The figure which shows the mode of the slider operation in the interface of FIG. The figure which shows the mode of the end point setting in the interface of FIG. The figure which shows the detailed structure of the interface of FIG. It is a figure for demonstrating range specification by the interface of FIG. 12, Comprising: The figure which shows a mode that an attention point is moved to a starting point Diagram showing the determination of the start point in range specification The figure which shows a mode that an attention point is moved to an end point in range specification Diagram showing end point determination in range specification The figure which shows the setting of the 2nd starting point in range specification The figure which shows after the setting of the 2nd starting point in range specification Diagram showing selection of second end point in range specification The figure which shows an example of the data content which memorize | stores the setting of how to superimpose Diagram showing an example of how to superimpose FIG. 10 is a diagram for explaining a preview presentation interface according to the third embodiment and shows a series of input video frames. Diagram showing one frame that is manually cut out The figure which shows the manual cutting result of the flame | frame of FIG. The figure which shows the production | generation example of the strobe synthetic | combination image containing the flame | frame of FIG. Figure showing an example of the preview presentation interface The figure which shows after correction by preview presentation interface Diagram for explaining the conventional chroma key method Diagram showing input video frame The figure which shows the video frame after duplicating three layers Diagram showing video frames with layers shifted in the time direction The figure which shows the image of the 1st frame of each layer of FIG. The figure which shows the strobe composite image which overlap | superposed the 1st frame of FIG. Flowchart of Strobe Image Composition Method According to Fourth Embodiment of the Present Invention Flowchart of Strobe Image Composition Method According to Fifth Embodiment of the Present Invention The figure which shows an example of the user interface used in order to make a user select a desired strobe synthetic | combination image. Flowchart of Strobe Image Composition Method According to Eighth Embodiment of the Invention The block diagram which shows schematic structure of the flash image synthesizing | combining apparatus which concerns on 9th Embodiment of this invention. Flowchart of Strobe Image Composition Method According to Sixth Embodiment of the Present Invention Flowchart of Strobe Image Composition Method According to Seventh Embodiment of the Present Invention The figure which shows an example of the process which extracts an area | region automatically from the image for correct data A flowchart of a strobe image composition method according to the tenth embodiment of the present invention. The flowchart of the strobe image composition method according to the eleventh embodiment of the present invention. Figure showing an example of feature point tracking results The figure which shows a mode that an exercise | movement pattern determination process is performed The flowchart which shows the flow of a process of the strobe image composition method concerning 12th Embodiment of this invention.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 ... Display apparatus, 2 ... Central processing unit, 3 ... Input device, 4 ... Main memory, 5 ... Strobe image composition program, 6, 7 ... Data

Claims (3)

In an image composition method for generating a strobe composite image by superimposing a plurality of frames of moving images,
Inputting feature points of the subject;
A feature point tracking step of tracking a movement trajectory of the feature point in the strobe composite image to obtain a trajectory pattern;
And a step of displaying the obtained pattern of the trajectory. In an image composition device that generates a strobe composite image by superimposing a plurality of frames of moving images,
An input means for inputting feature points of the subject;
Feature point tracking means for tracking the movement trajectory of the feature point of the subject in the strobe composite image to obtain a trajectory pattern;
An image synthesizing apparatus comprising display means for displaying the obtained locus pattern. In an image composition program that creates a strobe composite image by superimposing multiple frames of moving images,
The procedure for inputting the feature points of the subject,
A feature point tracking procedure for tracking a movement trajectory of the feature point of the subject in the strobe composite image to obtain a trajectory pattern;
An image synthesis program for causing a computer to execute a trajectory pattern display procedure for displaying the obtained trajectory pattern.

Images