JP2856299B2 - METHOD AND APPARATUS FOR PREPARING A PREVIEW IMAGE - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image preview method for grasping the contents of an image in a short time. In particular, it relates to a video content compression / editing and display processing method for grasping the content of a video in a short time, and more specifically, for example, a trailer for advertising a movie, a catalog creation for grasping a movie content, The present invention relates to a video preview method for summarizing video information such as grasping the outline of a self-recorded broadcast video and searching.

[0002]

2. Description of the Related Art It is possible to quickly grasp the contents of a sentence by skipping kanji and the like to some extent. In addition, sentences are easier to summarize than videos.
Further, the character information is a symbol, and is likely to be subjected to a search process on a computer or the like. On the other hand, it is difficult to watch a moving image quickly. For this reason, the conventional quick-view method has been limited to fast-forwarding in which the playback speed is increased to shorten the playback time, or variable-speed playback in which the viewer can change the playback speed to some extent. Also, images are difficult to understand by machine, and difficult to summarize. Since it is difficult to attach a search key, search processing cannot be performed as easily as characters. In the case of a moving image, a method of attaching a signal serving as a landmark to a certain frame in advance and searching for the signal based on the signal is performed. In a video disk or the like, a video is divided into several parts in advance in time, and a searchable mark is attached to the divided part for convenience of search. In addition, some VTRs automatically record a signal for searching on a video tape at the start of recording, thereby improving the efficiency of searching for a recording start point.

[0003]

The quick-forwarding of a moving image by fast-forwarding is easy to realize and the operation of the viewer is easy. However, since the moving image is displayed at a high speed regardless of the video content, the content may be overlooked. In addition, in the case of a moving image having a sharp movement, the display becomes very difficult to see, so that the observer tends to get tired and it is not practical to watch for a long time. In order to solve the above problem, the playback speed must be reduced, which is inconsistent with the original fast forward operation. Further, for these reasons, the reproduction speed cannot be increased so much, and it is difficult to reduce the display time.

[0004] In the case of variable-speed reproduction, the observer can adjust the reproduction speed according to the intensity of the movement and the degree of interest, but all adjustments must be performed by humans.
This operation is cumbersome at first, and the observer must determine the reproduction speed by looking at the display screen, so that the adjustment ability depends on the recognition ability of the observer. If the observer overlooks, there is no way to save it. To eliminate oversight,
The only option is to slow down or concentrate significantly, which increases the time required for quick viewing and fatigues the observer. Further, since the adjustment method is determined after the observer recognizes and the adjustment is actually performed, the adjustment always has a delay, and it is necessary to frequently perform reverse reproduction. This increases the time at which the watch can be seen, and can be mentally irritating to the observer.

[0005] In addition, the method of adding a signal for retrieval to a moving image in advance requires that the signal be added by a human, and the processing is cumbersome and time-consuming. Therefore, this method is hardly practical except when a large amount of the same moving image such as a VTR or an LD (laser disk) is created for a movie or the like, or when a specific moving image that is frequently viewed personally is created. There is also a problem that the place where the signal is attached depends on the subjectivity of the person attaching the signal.

As described above, when it is desired to grasp the contents of a moving image in a short time, it is difficult to grasp the contents in a short time by the conventional method. A method for automatically controlling and displaying the speed is proposed in Japanese Patent Application No. Hei.
No. 20519. However, the present inventor has conceived of further improvements and has accomplished the present invention. That is, an object of the present invention is to focus on the semantic role of each shot corresponding to one sentence in a sentence in a video, and to improve a quick video creation such that the quick understanding does not significantly lower the content understanding of the video. Is to provide a way.

[0007]

According to the present invention, there is provided a method for creating a preview image, comprising a series of frames connected in the order of display timing, wherein each shot divided by a cut point forms a continuous image. A readout method selecting step of selecting any one of a diagonal image readout method and a video search and readout method, comprising: When selected, a quick watch section that is a quick watch section, a quick watch time length that is a quick watch time length, and a quick watch frame position indicating a quick watch section in each shot of the quick watch section are set. Condition setting step to perform, and the quick time length for each shot in the quick section,
A preliminary frame extraction step of extracting from each shot a frame number corresponding to the allocated preliminary preview time, which is a frame number corresponding to the allocated preliminary preview time, and extracting the preliminary preview frame from each shot as a preliminary frame. A quick-view video editing step of connecting in accordance with the order of display timing and displaying on a display, and when a video search and read method is selected, a quick-view section that is a quick-view section and a time length of a quick-view A condition setting step of setting the quick watch time length, and assigning the quick watch time length to each shot of the quick watch section at a predetermined rate,
A preview frame extracting step of extracting a number of frames corresponding to the allocated shot preview time as a preview frame at equal intervals from each shot, and a preview video editing step of connecting the extracted preview frames in a display timing order and displaying the extracted preview frames on a display. Having. In addition, the flash image creation device of the present invention is configured by a series of frames connected in the order of display timing, and each shot divided by a cut point is
A quick-view video creating apparatus for quick-viewing a continuously formed image on a display, wherein the quick-view section is a section where a quick watch should be performed, a quick-view time length which is a time length for a quick watch, and a section where a quick watch should be performed. And an input device for setting a condition of a flash frame position indicating a portion to be flashed in each shot, and selecting any one of a video oblique reading method and a video searching and reading method, and a cut point of a stored cut point. A cut point storage device for outputting information, and each shot is detected based on cut point information from the cut point storage device, and when the image oblique reading method is selected from the input device, a quick preview from the input device is performed. A time length is assigned to each shot in the quick preview section at a predetermined rate, and the number of frames corresponding to the assigned shot quick preview time is a frame number of the quick preview frame (for example, The frame at the head or to select the position of the tail) of each shot was extracted from each shot as Hayami frame, connects the extracted Hayami frame in order of display timing, is displayed on the display unit,
When the video search and reading method is selected from the input device, the preview time length from the input device is assigned to each shot in the preview section at a predetermined rate, and the number of frames corresponding to the assigned shot preview time is assigned. A quick control device that extracts frames as quick frames at equal intervals from each shot, connects the extracted quick frames in the order of display timing, and displays the frames on a display.

[0008]

Therefore, according to the present invention, a summary video is created by automatically editing a representative video of each shot in accordance with a quick watching time designated by a viewer using information of each shot of the target video. At this time, the following editing is possible according to the request of the viewer.

When it is desired to grasp the outline of an image by an image oblique reading method corresponding to oblique reading when reading a book, a frame of a temporally continuous representative section is taken out from each shot, and it is played back in the standard order of shots. Play at speed. The length of the target section is automatically determined by calculation from the quick viewing time length specified by the viewer and the total number of shots of the target video.

[0010] Further, when it is desired to perform a video search and readout method, which is a quick look without any omission, such as searching for a specific operation or scene, the display time per shot is set, for example, to be uniform, and the display time is set within that time. The representative screens that can be selected are temporally evenly selected from the shots, and are reproduced in chronological order. At this time, the display time per shot is automatically determined by calculation from the quick watching time specified by the viewer and the total number of shots of the target video.

[0011]

Next, embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram showing an embodiment of a flash image creation device according to the present invention. The video reproduction device 4 outputs a video signal based on the control signal 8. In addition, high-speed video output is enabled according to a desired degree of quick viewing. The display device 1 displays a video based on a video signal from the video playback device 4. The cut point detector 9 detects a cut point which is usually composed of a plurality of frames and indicates the start and end of a shot corresponding to one sentence in a sentence, based on the video signal and the control signal 10 from the video reproducing device 4. . The operation of the cut detector 9 is described in detail in Japanese Patent Application No. 3-192997 filed by the present applicant. , The cut is automatically detected.

The cut point storage device 5 includes a cut point detector 9
Is stored in association with the frame number. Based on the operator's instruction, the input device 2 corresponds to a flash section which is a section to be flashed, a flash time length which is a time length to be flashed, and a portion to be flashed in each shot of the section to be flashed. And a control signal 6 for selecting one of the image oblique reading method and the image searching and reading method.

The image oblique reading method referred to here corresponds to oblique reading or skipping reading in the case of reading a book, and is a case where it is desired to grasp the outline in a short time. In this case, the visibility is prioritized, and there may be some oversight after the approximate line is finished. In addition, the video search and read method is equivalent to quick reading and search reading of a book, and is a case where reading must be performed in a short time but with attention, and more specifically,
This is the case when trying to get as much information as possible in a short time, or to look for a specific scene or event without overlooking it. In this case, it is required that there is no oversight even if it becomes somewhat difficult to see.

When inputting a control signal 6 for instructing the above-mentioned conditions and a reading method from the input device 2, the quick viewing control device 3 performs the following processing and control (1) or (2).

(1) A case where the image oblique reading method is selected. The total number of shots of the video to be viewed at a glance is counted from the information stored in the cut point storage device 5. Allocate to all shots of the number that counts the quick time length. A frame corresponding to a designated portion (for example, the end portion of a shot in the case of a movie or the beginning portion of a shot in the case of news) corresponding to the allocated time is continuously selected.
The control signal 8 including the frame number of the frame to be displayed is output to the video playback device 4 so that the frames of the selected shots are connected in the order of display timing and displayed on the display device 1. (2) When the video search and reading method is selected. The total number of shots of the video to be viewed at a glance is counted from the information stored in the cut point storage device 5. The quick look time length is assigned to all shots counted. The number of frames to be viewed at a glance in each shot corresponding to the allocated time is calculated. Frames at specified intervals (for example, equal intervals) in each shot are extracted by the calculated number of frames. The control signal 8 including the frame number of the frame to be displayed is output to the video reproducing device 4 so that the extracted frames of the shots are connected in the order of display timing and displayed on the display device 1.

If the cut point information of the video to be viewed at a glance has been detected in advance, the cut point information is directly supplied to the cut point storage device 5, so that the flash video creation device is as shown in FIG. Configuration.

Next, the operation of the embodiment of FIG. 1 will be described with reference to FIG. By operating the input device 2 by the operator, starting and ending positions of the image to be Hayami subject, the time length to be used in Hayami Hayami time length T _b, portions to be the Hayami in each shot, reading for Hayami A method (image oblique reading method or image searching and reading method) is input (step S1).

The quick control device 3 detects a video to be flashed from the start position and the end position of the video, outputs the detected video from the video reproducing device 4 to the cut point detector 9, and outputs the detected video to the cut point detector 9. The cut point detector 9 is instructed to detect a cut point, receives the instruction, detects a cut point of the video from the video playback device 4, and stores information of the detected cut point in the cut point storage device 5. reads the information stored cut-point Hayami controller 3 counts the number of shots N _s with dividing the image in each shot (step S2).

[0019] Hayami control unit 3, an interrupt stop is used Hayami in each shot from Hayami time length T _b and the number of shots N _s unless indicated time _{T h (T h = T b} /
N _s ) is obtained (step S3), and the quick viewing method determines which of the image oblique reading method and the image searching and reading method is selected (step S4).

If it is determined in step S4 that the image oblique reading method has been selected, the quick control unit 3 determines whether or not to perform a quick preview in each shot (for example, the start portion at the beginning or the end portion at the end). The video reproducing apparatus 4 selects a certain predetermined number of continuous frames (step S5), connects them in the order of display timing, and reproduces them at the standard speed.
(Step S6). For example, "I would like to view a one-hour documentary with a six-minute video oblique reading method."
And an instruction to that effect is given to the input device 2. Further, when the documentary is that it was 600 shots, the compression ratio is 10, the display time T _h of each shot is 0.6 seconds. Since the standard playback speed is 30 frames / sec, 0.6 × 30 = 1 from each shot
Eight frames are selected and further connected to form a 6-minute video, which is displayed on the display device 1.

[0021] Hayami control unit 3, when it is determined that the read process looking image is selected in step S4, the time T _h viewable frame number N _f (N _f in each shot
= _Th / number of display frames per second) (step S)
7), indicating the calculated number of frames (for example,
A predetermined number of frames are selected (step S8), and the video reproducing apparatus 4 is instructed to connect the frames in the order of the display timing and to reproduce the frames at the standard speed (step S9). For example, as in the above example, it is assumed that there is a request “I want to watch a one-hour moving image by a video search and read method for six minutes”, and an instruction to that effect is given to the input device 2. In addition, the documentary is when it was 600 shot, the display time T _h of each shot is 0.6 seconds. Therefore, a shot is 6 seconds (the standard playback speed is 30 frames / second, so a 6-second shot is 6 × 30
= 180 frames), the shot includes 180 frames. Also, since the display time _Th is 0.6 seconds, 18 images (0.6
(× 30 = 18 frames). Therefore,
Selection of a frame in this shot is made every 10 frames (180/18 = 10). The selected frame is further connected to be a 6-minute video and displayed on the display device 1. Interruption for stopping reproduction, rewinding, and changing T _b , quick view method, etc. is executed (step S1).
0), when it is determined that the stop of the process being executed has been instructed (step S11), the process returns to step S1 to restart the process.

In the above-described embodiment, the automatic editing is performed according to the quick playback time length, which is a preset reproduction time, and the reading method. In this case, the automatic editing is performed.
It may be performed sequentially in real time or may be performed in advance. For example, in the case of a medium such as a video tape that cannot display a designated frame at high speed, the medium is edited in advance and reproduced later. On the other hand, in the case of a video medium, for example, a video disk, which can access an arbitrary frame at high speed, it can be edited and displayed one by one. In this case, the viewer can change the reproduction time and the reading method at any time.

[0023]

As described above, according to the present invention, when the image diagonal reading method is selected, all the shots are not lost, and a part of the shots is reproduced at the standard speed, so that it is easy to see.
In addition, it is possible to quickly see the rough lines, and it is possible to answer a request for a quick overview of the video for a long time to grasp the outline. If the video search and reading method is selected,
You can see short and long shots without omission, and you can watch the shots as completely as possible from start to finish, and you can respond to requests to see exactly what kind of video was or to search for something.

As described above, by selecting and displaying the representative screen for each shot, the oblique reading and the search reading like a book can be automatically processed even in the video. Depending on the time available for quick viewing, the quick viewing video can be automatically edited, and it is also possible to select a method that is suitable for each, whether you look at the overview or look for something in detail, and more flexible Handling becomes possible.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an embodiment of a flash image creation device according to the present invention.

FIG. 2 is a block diagram showing another embodiment of the preview image creation device of the present invention.

FIG. 3 is a flowchart showing the operation of the embodiment of FIG. 1;

[Explanation of symbols]

 Reference Signs List 1 display device 2 input device 3 quick control device 4 video playback device 5 cut point storage device 6, 8, 10 control signal 7 cut point information (frame number) 9 cut point detector S1 to S11 Step

Continued on the front page (58) Fields surveyed (Int.Cl. ⁶ , DB name) H04N 5/7826 G11B 27/02 H04N 5/783 H04N 5/93

Claims (5)

(57) [Claims] 1. A quick preview of an image formed by a series of frames connected in the order of display timing, wherein each shot divided by a cut point is formed continuously on a display of a quick preview video creating apparatus. In the video creation method, a flash section, which is a section to be flashed, a flash duration, which is a time length to be flashed, and a flash frame position indicating a flashed portion in each shot of the section to be flashed are set. A condition setting step, and a preview time length is assigned to each shot in the preview section at a predetermined rate, and the number of frames corresponding to the assigned shot preview time is defined as a preview frame. The step of extracting the quick frames extracted from the shots and the extracted quick frames are connected in the order of the display timing and displayed on the display. A quick video editing step. 2. A quick preview of a video composed of a series of frames connected in the order of display timing, wherein each shot divided by a cut point is continuously formed on a display of a quick preview video creating apparatus. In the video creation method, a condition setting step for setting a quick-stop section which is a quick-stop section and a quick-stop time duration which is a quick-stop time section; A quick frame frame extracting step of extracting frames at the same rate as the number of frames corresponding to the assigned quick preview time from each shot at equal time intervals, and connecting the extracted quick preview frames in the order of display timing;
A preview video editing step of displaying the preview video on a display. 3. A quick preview of an image formed by a series of frames connected in the order of display timing, wherein each shot divided by a cut point is formed continuously on a display of a quick preview video creating apparatus. In the video creation method, a readout method selection step of selecting one of a video diagonal readout method and a video search and readout method, and, when the video diagonal readout method is selected, a quick look section which is a section to be flashed A condition setting step of setting a quick watch time length that is a quick watch time, a quick watch frame position indicating a portion to be quick watch in each shot of the quick watch section, and a quick watch time length of each shot of the quick watch section And the number of frames corresponding to the assigned shot preview time and the frame at the preview frame position. Frame extracting step for extracting a frame as a quick frame from each shot, and a quick video editing step of connecting the extracted quick frames in the order of display timing and displaying the same on a display, and a video searching and reading method. When is selected, a quick-fit section that is a quick-watch section, and a condition setting step of setting a quick-watch time length that is a quick-watch time length, and a quick-watch time length for each shot of the quick-watch section. Allocating at a predetermined rate, extracting a number of frames corresponding to the allocated shot preview time as preview frames at equal intervals from each shot, and connecting the extracted preview frames in display timing order,
A preview video editing step of displaying the preview video on a display. 4. The quick preview image creating method according to claim 1, wherein the quick preview frame extracting step assigns a quick preview time length equally to each shot in the quick preview section. 5. A quick-view video creating apparatus which comprises a series of frames connected in the order of display timing, wherein each shot divided by a cut point is viewed on a display at a quick glance at an image formed continuously. Obliquely read video by conditionally setting a flash section, which is a section to be flashed, a flashing time length, which is a time length to be flashed, and a flash frame position indicating a flashed portion in each shot of the section to be flashed. An input device for selecting any one of a method and a video search and readout method; a cut point storage device for outputting stored cut point information; and a cut point information from the cut point storage device. In addition to detecting each shot, if the video oblique reading method is selected from the input device, the preview time length from the input device is set for each shot in the preview section. A frame number assigned at a predetermined rate and corresponding to the assigned shot preview time, the frame at the preview frame position is extracted as a preview frame from each shot, and the extracted preview frames are connected in the order of the display timing. When the video search and read method is selected from the input device, the preview time length from the input device is assigned to each shot in the preview section at a predetermined rate,
A preview control device for extracting frames of the number of frames corresponding to the assigned preview time as a preview frame at equal intervals from each shot, connecting the extracted preview frames in display timing order, and displaying the frames on a display. Quick-view video creation device.