JP2012156686A - Search method, search device, and moving image editing device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To easily and efficiently perform a work to search for a target candidate from multiple candidates.SOLUTION: A moving image editing device 1 searches for a division position in dividing a moving image as follows by the function of a moving image editing application program. First, a whole moving image or a part of it is set as the population data of a search target, N frame images are extracted from the data in accordance with a predetermined extraction rule such as equal interval, and then, an arrayal list R0 is displayed, where the thumbnails of the extracted frame images are arrayed by time sequence. When a touch input is performed at a prescribed position of the arrayal list R0, a moving image in a range between the thumbnails immediately before and after the input position is set as the new population data. Then, till the displaying of the thumbnail of the target frame image, three steps are repeated, i.e., narrowing the population by the touch input, extracting the N frame images, and displaying the list of the extracted thumbnails.

Description

  The present invention relates to a search method and a search apparatus for searching for a target candidate from among a large number of candidates, and more particularly to efficiently search for a target frame image from data including a very large number of frame images such as moving images. The present invention relates to a search method and a search apparatus suitable for the above, and a video editing apparatus suitable for efficiently searching for a target frame image from a video to be edited by using this search method.

  In order to perform editing processing such as processing to divide a video at a specific location, processing to cut out a specific range, or processing to delete a video such as recording data of a TV program or shooting data taken with a home video, It is necessary to accurately detect the start position and end point of the corresponding position or range from the moving image and specify them as the division position or the cutout position.

  Conventionally, as a method for accurately specifying the desired position and range in such a movie, the movie is played back in chronological order, and playback is stopped at that point when the movie is played back close to the target location. Then, a method is used in which the frame images before and after the frame image are confirmed one by one by frame advance and the frame at the target position or the frame that is the start point or end point of the target range is designated.

  However, with this method, when the distance from the position where playback was stopped to the target frame is far, it takes time and labor to perform frame-by-frame advancement by the frame advance button. In addition, since such a frame advance operation takes a long time from when the frame advance button is pressed to when the frame image is switched, the frame advance button may be pushed too much and a frame to be designated may be passed. Another problem is that if a frame to be designated is passed, the target is lost, and it takes more time.

  As another method for accurately specifying a desired position or range in a moving image, a predetermined number of frame images centered on the playback position specified by the scroll bar are displayed on the screen in time series together with the scroll bar. By displaying the frame list arranged in a line and scrolling the frame list by operating the scroll bar, the frame images are sequentially confirmed along the time series to detect the target position or the start point or end point of the target range. Method is used.

  However, with this method, the number of frames that can be displayed on one screen is at most about ten. In contrast, normal moving image data is composed of 20 to 30 frame images per second, and the number of frames that can be displayed on one screen is about 0.5 seconds or less as a moving image playback time. It must be. Therefore, in the case of a moving image having a huge number of frame images, it takes time to scroll to the target frame. In addition, the operation of scrolling the image on the frame list at an appropriate speed is difficult because it is a fine operation, and the work load is large. Furthermore, if the user is unaccustomed to fine work and the scroll speed is too high, the target frame image may be missed, and the target may be lost due to scrolling too much.

  As described above, in the conventional method, it is necessary to sequentially switch the frame images in time series or scroll the frame list until the target frame is reached. It's not efficient. Therefore, a method for searching for a target location in a moving image without performing such frame advance display or scroll display has been proposed. Patent Document 1 discloses this kind of search method.

  In the scene search method of Patent Document 1, a scene (frame image) is extracted from a moving image for each unit time set in advance by a user. Then, the thumbnails of each scene are displayed side by side in a matrix on the screen. When the user selects any one of the thumbnails on the screen, the thumbnails of the scenes extracted at finer time intervals centering on the scene corresponding to the selected thumbnail are displayed. In the method of Patent Document 1, by repeating such a scene extraction step and a scene selection step, the moving image is narrowed down until the scene extraction interval becomes the shortest unit time among preset unit times. It is.

JP 2006-186617 A

  In the method of Patent Document 1, a unit time preset by the user is used as a time interval for extracting a scene (frame image) from a moving image in the first scene extraction step. For this reason, the number of scenes to be extracted differs according to the length of the moving image, and a large number of scenes are extracted from the long moving image data, and only a small number of scenes are extracted from the short moving image data. Therefore, in the case of a short video, depending on the unit time setting, the number of initial scene extractions becomes too small, and the initial scene selection does not make much sense and becomes inefficient. is there.

  In the method of Patent Document 1, the number of extracted scenes (frame images) varies depending on the length of the moving image as described above, while the layout for displaying the thumbnails of the extracted scenes is 9 per screen. The number of displays is fixed, such as a scene or 16 scenes. For this reason, in the case of a long movie, it is not possible to list thumbnails of all scenes on one screen in the initial scene selection process, and the screen must be switched to check all the thumbnails. . In addition, since thumbnails are displayed in a matrix such as 3 rows × 3 columns, 4 rows × 4 columns, etc., the time series order of the displayed thumbnails is difficult to understand, and whether or not the target scene is the target scene. There is also a problem that it is difficult to judge. Therefore, it cannot be said that the method of Patent Document 1 can perform scene selection sufficiently efficiently.

  Further, in the method of Patent Document 1, when the scene extraction process and the scene selection process are repeated to narrow down the scene hierarchically, the time interval in the second and subsequent scene extractions is set to n / (A × b). (N: time interval at the time of first scene extraction, A: number of times of extraction (second is set to 1), b: specified number). In other words, in this method, the extraction interval becomes finer every time the number of extractions is repeated, but the extraction interval is only linearly reduced. Therefore, when considering a huge number of candidate data such as frame images of moving images, even a moving image of about 30 minutes includes tens of thousands of frame images. If an attempt is made to determine the scene switching position after confirmation, it is necessary to repeat the extraction process and selection process at least several hundred times or more. Therefore, it cannot be said that the search operation can be performed sufficiently efficiently.

  In addition, since the method of Patent Document 1 selects only one time point in a moving image and extracts a scene centering on that time point, when searching for a specific range, two points of a start point and an end point are selected. You have to search twice. Even in the conventional method, there are those that can select an arbitrary range in the video, but in the end, in order to accurately determine the start point and end point, the work of narrowing down each of the start point and end point is divided into two times. It takes time and effort.

  In view of the above-described problems, an object of the present invention is to propose a search method and a search apparatus that are suitable for easily and efficiently performing a task of searching for a target candidate from among a large number of candidates. is there. Another object of the present invention is to propose a moving image editing apparatus that can easily and efficiently search a target location from a moving image to be edited by using this search method.

In order to solve the above problems, the search method of the present invention provides:
Based on an input for designating a search target or instructing start of search, the control means performs a process of setting a series of candidate data that can be arranged according to a predetermined sequence standard as a population to be searched; and
An extraction step in which the control means performs a process of extracting a preset number of candidate data according to a preset extraction rule from the population;
A list display step in which the control means generates a sequence display in which the display indicating the extracted candidate data is arranged in a line in accordance with the sequence criteria, and performs a process of displaying the sequence display on the display means as a list;
The control unit waits for a selection input for selecting a part of the array display, and when the selection input is performed, which part of the array display is selected based on the selection input. And performing a process of setting a portion of the population corresponding to the selected portion as a new population,
The control means, when the population is set, until an input for ending or interrupting the search is performed, or until it is determined that the population does not need to perform the extraction step , The extraction process, the list display process, and the narrowing-down process are repeated.

  In the present invention, when a population is set as described above, a predetermined number of candidate data is automatically extracted from the list and displayed as a list, so that the candidate data can be displayed one by one or finely scrolled. The candidate data can be confirmed and selected without any operation or screen switching. Therefore, it is possible to easily search for candidate data. In addition, since a certain number of candidate data is always extracted regardless of the amount of candidate data as a population, the number of candidate data to be displayed in a list is not halfway, and the search can be efficiently performed. Further, since the range of the population is narrowed down exponentially each time extraction and selection of candidate data is repeated, even when narrowing down a large number or a large range of candidate data, it is possible to proceed to the final stage of the search process with a small number of selections. The target candidate data can be reached with a small number of selections. Therefore, efficient search is possible. In addition, since the number of candidate data to be displayed is relatively small even when the search process proceeds to the final stage, the processing load for displaying the array indicating the extracted candidate data is small.

  Here, when the control unit determines that the population includes only the discrete candidate data equal to or less than the preset number, the control unit stops the repetition of the three steps and configures the population. A process of displaying all candidate data in a line and displaying the list on the display means is performed. Thus, in the case of a search for discrete candidate data instead of continuous candidate data, only the candidate data equal to or less than the number set in advance at the final stage of the search process remains. Therefore, in this case, it is determined that the extraction process is not necessary, and the candidate data to be targeted can be determined by displaying all candidate data constituting the remaining population.

  Further, in this case, the control means, in the extraction step, when dividing the array in which all the candidate data constituting the population are arranged in a line according to the sequence criterion into a predetermined number at a preset division ratio The preset number of candidate data can be extracted so as to include candidate data corresponding to each division position. In this way, appropriate candidate data can be extracted according to the contents of the candidate data and the search purpose.

  Alternatively, the control means, in the extraction step, as the preset number of candidate data, the candidate data at the beginning and the end of the array in which all candidate data constituting the population are arranged in a line according to the sequence criterion, In addition, candidate data corresponding to each division position when the array is equally divided into a predetermined number may be extracted. In this way, candidate data can be extracted uniformly while covering the entire range of the population.

  Here, in the list display step, the control means is arranged so that an array direction of the array display to be displayed in a list is linearly arranged in any one of vertical, horizontal, and diagonal directions, or an arc shape or an annular shape. The arrangement direction can be arranged in a row. As described above, according to the present invention, appropriate array display can be performed according to the contents of the candidate data, the array reference, the layout of the display screen, and the like. For example, in the case of data arranged in time series, the context is easily understood by arranging the data in a straight line. In addition, when an image group in which the hue is continuously changed is used as a candidate, display along the hue ring can be performed by arranging the images in an annular shape.

  In the present invention, when the series of candidate data is a moving image composed of a plurality of frame images that are sequentially switched and displayed at predetermined time intervals, the control means sets the population in the extraction step. A process of extracting the preset number of frame images from the preset number of extraction positions set to be distributed over the entire range of the moving image, and in the list display step, the extraction is performed. A process of generating an array display in which the thumbnails of the frame images extracted in the process are arranged in a line according to the display order of the frame images and displaying the array display in a list on one screen is performed. When it is determined that a part of the array display has been selected, the range of the moving image corresponding to the selected part is set as a new population. When the population is set, it is determined that an input for terminating or interrupting the search is performed, or the number of frame images constituting the population is equal to or less than the preset number. Until the control unit determines that the number of frame images constituting the population is equal to or less than the preset number, the extraction step, the list display step, and the narrowing step are repeated. It can be configured to perform a process of displaying a list of array displays in which thumbnails of all frame images constituting the population are arranged in a line.

  As described above, when the present invention is applied to search for frame images constituting a moving image, the range of the moving image is narrowed down without performing fine and troublesome operations such as frame list scrolling, frame advancement, and screen switching. be able to. In particular, every time you specify a range for an array of thumbnails, the range of the video can be narrowed exponentially, so even if you search from a video that contains a large number of frame images, you can specify the range a few times. To reach the target frame image. Therefore, efficient search is possible. In addition, since the number of thumbnails to be displayed before the search is completed is not so large, the processing load for generating and displaying thumbnails is small.

  Here, when the control unit determines that the portion between the two adjacent thumbnails in the array display is selected in the narrowing-down process, one of the two frame images corresponding to the two adjacent thumbnails is selected. Can be configured to set the range of the moving image having the starting point and the other end point as the new population. In this way, it is possible to narrow down the range of moving images (frame image group) constituting the population at a magnification according to the number of extracted frame images. Further, since it is only necessary to select an arbitrary position between thumbnails, an operation for specifying a range is simple.

  Alternatively, when the control means determines that a section from any thumbnail to another thumbnail in the array display is selected in the narrowing-down step, the thumbnail located at the head of the selected section and its thumbnail Two frame images corresponding to the thumbnail located immediately before are set as the start point and end point of the first range in the moving image, and correspond to the thumbnail located at the end of the selected section and the thumbnail located immediately after it. Two frame images to be set as a start point and an end point of a second range in the moving image, the first and second ranges are set as the new population, and the first and second ranges are set. In the extraction step immediately after setting a range as a new population, each of the first and second ranges It can be configured to a have been set the number of extraction positions set in advance so as to disperse by constant, extracting the frame image number with the preset. In this way, when determining a predetermined range such as a portion to be cut out of a moving image or a portion to be deleted, the start point and the end point can be narrowed down at a time. In addition, since most of the range between the start point and the end point is excluded from the search target, the range of the moving image can be efficiently narrowed down to two places near the start point and the end point. Therefore, the search can be performed efficiently regardless of the length of the target moving image range.

  Next, in the present invention, in the case where the series of candidate data is data of a flight group of transportation facilities, the control means presets a departure / arrival time range of the flight group in the extraction step. It can be set as the structure which extracts the data of the said preset number of operation flights so that the data of each operation flight nearest to each division | segmentation time divided | segmented into the predetermined number by the division | segmentation ratio may be included. As described above, the present invention can be applied to a case where a flight is searched from data such as a timetable, and can be used as a method for easily and efficiently searching for a flight.

  In the present invention, when the series of candidate data is data of an image group corresponding to each value of the feature amount when a predetermined feature amount in the image is changed within a predetermined range, the control is performed. In the extracting step, the presetting is performed so as to include data of each image corresponding to each division point when the change range of the feature amount in the image group is divided into a predetermined number at a preset division ratio. It is also possible to extract the data of the number of images. As described above, the present invention can be applied to the case where feature quantities such as hue and lightness in an image are continuously changed, and can exemplify a predetermined number of images corresponding to each value in the change range. Therefore, the target feature value can be easily and efficiently determined by comparing and examining the exemplified thumbnails.

Next, the present invention is a search device for searching for target candidate data from a population to be searched by the above search method,
Display means for displaying a list of array displays in which a predetermined number of displays are arranged in a line;
Input means for receiving a selection input for selecting a part of the array display;
Storage means for storing a program for executing the search method;
And a control unit that executes processing for performing each step in the search method based on a program read from the storage unit.

Further, the present invention is a moving image editing apparatus for searching for a frame image of an edited portion in a moving image by the above search method,
Display means for displaying a list of array displays in which a predetermined number of thumbnails of frame images included in the video data are arranged in a line on a single screen;
Input means for receiving a selection input for selecting a part of the array display;
Storage means for storing a program for executing the search method;
And a control unit that executes processing for performing each step in the search method based on a program read from the storage unit.

  Here, in each of the above devices, when the display unit is a liquid crystal display panel and the input unit is a touch panel installed on the surface of the liquid crystal display panel, the control unit is When an operation of touching or tracing a part of the array display displayed in a list on the liquid crystal display panel is performed, the touched portion or the portion of the population corresponding to the traced portion (candidate data is the moving image data). In this case, the range of the moving image) can be set to a new population. Thus, when the present invention is applied to a search operation in an apparatus having a touch panel type input display panel, the operability at the time of search is good. For example, it is suitable for search work on a tablet computer or the like.

  According to the present invention, when a population is set, a predetermined number of candidate data are automatically extracted from the list and displayed as a list, so that fine operations such as displaying candidate data one by one or finely scrolling In addition, candidate data can be confirmed and selected without any screen switching. Therefore, it is possible to easily search for candidate data. In addition, since a certain number of candidate data is always extracted regardless of the amount of candidate data as a population, the number of candidate data to be displayed in a list is not halfway, and the search can be efficiently performed. Further, since the range of the population is narrowed down exponentially each time extraction and selection of candidate data is repeated, even when narrowing down a large number or a large range of candidate data, it is possible to proceed to the final stage of the search process with a small number of selections. The target candidate data can be reached with a small number of selections. Therefore, efficient search is possible. In addition, since the number of candidate data to be displayed is relatively small even when the search process proceeds to the final stage, the processing load for displaying the array indicating the extracted candidate data is small.

It is explanatory drawing of the moving image editing apparatus which is one Embodiment of the search device of this invention. It is a block diagram which shows schematic structure of the control system of a moving image editing apparatus. It is explanatory drawing which shows the search method of the frame image which shows the division | segmentation position in the case of dividing | segmenting a moving image at a predetermined position. It is a flowchart which shows the flow of a process at the time of performing the search method of FIG. It is explanatory drawing which shows an example of the application screen in list display. It is explanatory drawing which shows the search method of the start point and end point of the said range when deleting or cutting out a predetermined range from a moving image. It is a flowchart which shows the flow of a process at the time of performing the search method of FIG. It is a display example of an arrangement list when performing hue adjustment.

(Embodiment 1)
The search method and search apparatus (frame image search method and video editing apparatus in video editing) according to the first embodiment to which the present invention is applied will be described below with reference to FIGS.

(Device configuration)
FIG. 1 is an explanatory diagram of a moving image editing apparatus which is an embodiment of a search apparatus of the present invention. As shown in this figure, a moving image editing apparatus 1 (search apparatus) is a tablet computer having a flat outer shape, and a rectangular input display panel 2 is provided on one surface thereof. The input display panel 2 is configured to allow display and touch input by installing a touch panel 4 (input means) on the surface of the liquid crystal display 3 (display means). In addition, various operation switches including a power switch, connectors for connecting a power cable and various connection cables, and the like are provided on the edge of the casing 5 surrounding the input display panel 2 as necessary. On the back side of the input display panel 2, various hardware including a CPU 6 (see FIG. 2) and a circuit board on which various devices are mounted, a power supply device, and the like are mounted.

  FIG. 2 is a block diagram showing a schematic configuration of a control system of the moving image editing apparatus. The control unit 7 (control unit) of the moving image editing apparatus 1 includes an arithmetic processing unit such as the CPU 6 described above and a storage unit 8 (storage unit) that stores various programs and data necessary for control. The storage unit 8 includes a nonvolatile memory such as a flash ROM or an EEPROM, and a RAM that functions as a work area during processing by the CPU 6 or the like, and an auxiliary storage device such as an HDD or an SSD as necessary. Can also be used as part of it. Further, the control unit 7 is provided with a video controller 9 that generates drawing data of an image to be displayed in accordance with a drawing command from the CPU 6 and performs processing for displaying an image on the liquid crystal display 3 based on the drawing data. .

  An input signal from the touch panel 4 is input to the control unit 7 via an I / O controller (not shown), and an input signal from an input / output cable or an operation signal from various operation switches is input. The The control unit 7 reads necessary programs and data based on these input signals and operation signals, and controls each unit of the moving image editing apparatus 1. When the communication module for performing communication via wireless communication or a communication cable is mounted on the moving image editing apparatus 1, various control signals are transmitted between the control unit 7 and the external device via the communication module. And send and receive data.

  The form of the moving image editing apparatus 1 is not limited to the form as described above, and can be changed as appropriate. For example, an HDD or other device can be connected to the main body portion composed of the above-described tablet computer, and data can be obtained from these devices as necessary. In addition, the tablet type terminal having the input display panel 2 and the main body part having the control unit 7 and the storage unit 8 are configured separately, and are configured to be communicable via a connection cable or wireless communication. You may utilize the input means provided in the main-body part, and the input means provided in the terminal as needed. Moreover, it is good also as a structure which mounts the moving image edit application program which provides the function which performs the search method of this embodiment in the various computer apparatuses which have a display screen and an input means.

(Search method for frame images in video editing)
The storage unit 8 of the moving image editing apparatus 1 stores a moving image editing application program for performing moving image editing using the search method of the present embodiment. When an input for instructing execution of the moving image editing application program is performed from the touch panel 4 or the like, the control unit 7 reads the moving image editing application program based on this input. Based on this program, an input screen showing a list of moving image data stored in the storage unit 8 is displayed on the liquid crystal display 3. When a touch input for selecting any moving image data is performed on the touch panel 4, the control unit 7 determines the selected moving image data as an editing target, and performs a process of reading the moving image data. It is also possible to search the moving image data in the accessible external storage device by touch input and read the moving image data to be edited from the external storage device.

  The moving image editing application program provides a search function for searching for and specifying a specific portion in a moving image by each search method described below. The search target is assumed to be a specific frame image in a moving image or a frame image indicating a specific range. For example, a frame image indicating a division position for dividing a moving image at a scene switching position or the like is assumed. Further, when performing a cutout operation for cutting out a specific portion in a moving image or a deletion operation for deleting a specific portion, a frame image indicating the start point and the end point of the target portion is assumed. Note that the search target is not limited to these frame images, and any frame image can be set as the search target.

(1) Method of searching for division position in moving image division processing FIG. 3 is an explanatory diagram showing a method of searching for a frame image indicating a division position when a moving image is divided at a predetermined position, and FIG. 4 is a search method of FIG. It is a flowchart which shows the flow of the process at the time of performing. Hereinafter, a method of searching for a frame image indicating a division position will be described with reference to these drawings. When the touch input for performing the division work on the editing target moving image is performed from the menu screen or the like displayed on the liquid crystal display 3, the control unit 7 performs the processing of the flowchart of FIG. 4 based on the moving image editing application program. Start.

  First, in step S1 (initial setting step), the control unit 7 performs initial setting of a population to be searched. Specifically, the entire range from the beginning to the end of the moving image read as an editing target is set as a population. In step S1, as described above, based on the input for instructing the start of the division work, the entire moving image data currently being edited is automatically set as a population, and the moving image data is displayed to the user. A part may be specified, and only a part of the specified moving image may be set as the first population.

  Subsequently, the control unit 7 proceeds to step S2, and performs a comparison determination between the number of frame images constituting the population and the number N of frame images extracted in step S3 (a preset number). If the number of frame images in the population is greater than N (step S2: Yes), the process proceeds to step S3. On the other hand, when the number of frame images in the population is N or less (step S2: No), the control unit 7 determines that it is not necessary to narrow down the frame images, and proceeds to step S8 described later.

  Subsequently, the control unit 7 proceeds to step S3 (extraction process), and performs a process of extracting a preset number of extraction (N) frame images from a moving image that is a population according to a preset extraction rule. In this embodiment, when the user finds a frame image that does not know where in the moving image that is the population, a process of extracting the frame image equally from the entire range of the moving image is performed. Specifically, the top and bottom frame images of the moving image that is the population are extracted, and the positions of N-2 dividing points that are generated by equally dividing the entire moving image by N-1 are calculated. A frame image at each division point is extracted. Thereby, N frame images arranged at equal intervals over the entire range of the population can be extracted.

  FIG. 3A shows an example of the extraction result obtained in the first step S3. In the example shown in FIG. 3A, the moving image to be edited is composed of 30 frame images per second, and the total length is 30 minutes. The number N of frame image extractions is set to N = 8 in consideration of the thumbnail display size of each frame image in step S4 (list display step) described later. In this case, the number of frame images constituting the population is 30 × 60 × 30 = 54000. Therefore, a frame image of each division point when an array in which these are arranged in the display order is divided into seven (N-1 divisions) is extracted. As a result, as shown in FIG. 3A, the 7714th, 15428th, 23142th, 30857th, 38571th, and 46285th frame images are extracted as the frame images F02 to F07. Then, the first and 54000th frame images are extracted and added as the first and last frame images F01 and F08 of the moving image that is the population, so that a total of eight frame images F01 to F08 are extracted. The

  In this example, the frame image extraction interval from the frame image sequence of the population is 54000 ÷ 7 = 7714.28, and the division position does not become an integer. Therefore, the frame image immediately before or after the division position is By appropriately extracting the frame images corresponding to the division positions, the extraction intervals are made substantially equal. Further, in this example, instead of using the first frame image of the moving image as the frame image F01, the title image displayed before the moving image starts is handled as the 0th frame image, and this image is the first frame image. The frame image F01 may be used.

  Next, the control unit 7 proceeds to step S4 (list display step). In step S4, thumbnails are generated for each of the extracted N frame images (eight in the example shown in FIG. 3), and an array list R0 (array display: FIGS. 3 and 5) in which these thumbnails are arranged in a time series. The process of displaying a list on the input display panel 2 is performed. Then, the process proceeds to step S5 (part of the narrowing-down process) to enter a standby state where a touch input for the displayed array list R0 is performed or another input is performed.

  FIG. 5 is an explanatory diagram illustrating an example of an application screen during list display. As described above, when the moving image editing application program according to the present embodiment performs the operation of dividing the moving image, the generated thumbnail arrangement list R0 is displayed at the display position at the lower end of the input display panel 2, and the input display panel 2 is filled to the full width. Display with the length of. The moving image playback screen is displayed at the top. The moving image playback screen may be omitted, and only the thumbnail arrangement list R0 may be displayed on the screen, or may be displayed together with other moving image information. Moreover, the arrangement position and arrangement direction of thumbnails may be changed. For example, the thumbnails may be displayed in a vertical row or in a diagonal row, or may be displayed in an array shape such as an arc shape or an annular shape.

  The user examines the thumbnail arrangement list R0 displayed in step S4, and performs a touch input for selecting a range including the division position. For example, in the present embodiment, since the moving image set in the population is divided into seven, a range including a dividing position for dividing the moving image can be specified in units of the divided sections. For example, by performing touch input (selection input) to an arbitrary position between adjacent thumbnails, a section including this touch input position can be selected.

  When the touch input for selecting the moving image range is performed in the standby state in step S5, the control unit 7 proceeds to step S6 (part of the narrowing-down process). In step S6, the control unit 7 determines the selected portion based on the content of the touch input. For example, as indicated by an arrow A1 in FIG. 3A, when a touch input is performed on a position between two thumbnails indicating the frame images F05 and F06 on the array list R0, the control unit 7 performs the touch input. It is determined that a predetermined part including the position, for example, a section from frame images F05 to F06 corresponding to thumbnails before and after the touch input position is selected. And it progresses to step S7 (a part of narrowing-down process). In step S7, a process of resetting the selected range as a new population is performed. Then, returning to step S2, if the number of frame images constituting the population is larger than N (step S2: Yes), the process proceeds again to step S3.

  When the control unit 7 performs the process of step S3 again, N pieces arranged at equal intervals over the entire population as in the previous time with respect to the current population (range from the frame images F05 to F06). The process of extracting the frame image is performed. FIG. 3B shows an example of the extraction result obtained in the second step S3. In this case, since the frame images from the 30857th to the 38571th are the population, the frame image at each division point when the image is divided into 7 is extracted as in the first extraction step. As a result, as shown in FIG. 3B, the 31959th, 33061st, 34163th, 35265th, 36367th, and 37469th frame images are extracted as the frame images F12 to F17. Then, by adding the 30857th and 38571th frame images as the first and last frame images F11 and F18 of the population, a total of eight frame images F11 to F18 are extracted.

  Subsequently, the control unit 7 proceeds to the second step S4, generates the thumbnail arrangement list R1 of the frame images F11 to F18 and displays the thumbnail arrangement list R1 on the input display panel 2 as in the first time, and displays the second step S5. Proceed to enter the standby state. When touch input for selecting a range is performed, the process proceeds to step S6 for the second time, and the selected range is determined. For example, as shown by an arrow A2 in FIG. 3B, when a touch input is performed on a position between the frame images F12 and F13, the control unit 7 selects a range from the frame images F12 to F13. It is determined that Then, the process proceeds to step S7 for the second time, and a process for resetting the range from the selected frame images F12 to F13 as a new population is performed. And it returns to step S2 similarly to the last time. Similarly, the processes of S2 to S7 are repeated until the number of frame images in the population is N or less (step S2: No), or other input than the range designation is performed in step S5. As a result, the video is further narrowed down.

  When the list display step of step S4 is performed, the user confirms the thumbnail arrangement list Rn (n = 0, 1, 2,...) Displayed in this step. When it is determined that the thumbnail of the target frame image is included in the array list Rn, it is possible to input that the thumbnail is determined as a thumbnail indicating the division position of the moving image. . That is, in the standby state in step S5, the moving image editing apparatus 1 is in a state of accepting an input for determining the division position and ending the search in addition to the selection input for narrowing down the range of the moving image. When an input indicating that any one of the displayed thumbnails is selected and set as a division position is performed, the control unit 7 proceeds to step S10 and determines a frame image at the division position based on the input content. Then, it is determined that the search is finished, and the processing of this flowchart is finished. Then, the process proceeds to a process of confirming the search result or dividing the moving image based on the search result. It should be noted that another input can be made in step S5 to appropriately interrupt the search process or change the population.

  When the processing of steps S2 to S7 is repeated without determining the division position, the number of frame images included in the population decreases exponentially at a magnification of 1 / (N-1) every time the process is repeated. However, the frame image extraction interval is exponentially narrowed. FIG. 3C shows an example of the third extraction result. In the third time, the frame images F21 to F28 are extracted based on the touch input performed on the position between the frame images F12 and F13. As shown in FIG. 3, the extraction interval of the frame images F01 to F08 is every 7714 frames in the first extraction process, whereas it is every 1102 frames in the second time and every 157 frames in the third time. If the processing is continued in the same manner, the extraction interval is every 22 frames at the fourth time, and every three frames at the fifth time, so that a touch input for selecting a range is performed for the thumbnail display generated at the fifth time. Then, the population is narrowed down to four frame images.

  As described above, when the number of constituent elements of the population is equal to or less than the number N of frame image extractions (step S2: No), it is possible to display a list of thumbnails of all the frame images included in the population. Become. Therefore, in this case, it is determined that it is not necessary to perform further extraction steps, and the process proceeds to step S8. In step S8, thumbnails are generated for all frame images constituting the current population, and an array list in which these are arranged in a line in time series is displayed on the input display panel 2 as a list. Then, the process proceeds to step S9 and enters a standby state. In step S9, an input for determining the division position and other inputs are accepted. When a touch input for selecting any of the displayed thumbnails and setting the divided position is performed, the control unit 7 proceeds to step S10 and determines a frame image at the divided position based on the input content. Then, it is determined that the search is finished, and the processing of this flowchart is finished.

  As described above, in this search method, the number of frame images is narrowed down exponentially. Therefore, in the case of a 30 minute moving image (number of frame images: 54000), selection input for array display of frame images is performed at most 5 times. It is possible to proceed to the final stage of the frame image search only by performing (when N = 8). Therefore, the target frame image can be detected with a remarkably small amount of operation compared to the conventional method of confirming the frame image by scrolling or frame advance. In addition, it is not necessary to perform detailed operations such as frame advance and scroll, and it is only necessary to examine the thumbnails displayed in the list and touch-input the part close to the purpose, so the search work is easy and the burden on the user is also reduced. Few. Therefore, the search can be efficiently performed in a short time, and the usability is good. In addition, since there are not so many thumbnails to be displayed until the search is completed, there is an advantage that the processing load for generating thumbnails and displaying the arrays is small.

  Note that this search method can be applied not only to searching for a frame image indicating a division position for dividing a moving image, but also to searching for an arbitrary frame image. In this case, when an array display including thumbnails of the target frame image appears, input is performed to select this frame image, the search process is terminated, and the process is appropriately shifted to another process using the detected frame image. do it.

(2) Search method of start point and end point in processing to delete or cut out part of moving image FIG. 6 is an explanatory diagram showing a search method of the start point and end point of the range when a predetermined range is deleted or cut out from the moving image. FIG. 7 is a flowchart showing the flow of processing when the search method of FIG. 6 is executed. Hereinafter, a method of searching for frame images at the start point and end point of a predetermined range in a moving image will be described with reference to these drawings. When a touch input for performing a part of clipping operation or deletion operation is performed on a moving image to be edited from a menu screen or the like displayed on the liquid crystal display 3, the control unit 7 is based on the moving image editing application program. The process of the flowchart of FIG. 7 is started.

  First, the control unit 7 performs steps S11 to S14. Steps S11 to S14 are the same as steps S1 to S4 described above, and a description thereof will be omitted. FIG. 6A shows an example of the extraction result obtained in the first step S13. Similarly to step S3, a total of eight frame images G01 to G08 of the first or zeroth (title image), 7714th, 15428th, 23142th, 30857th, 38571th, 46285th, and 54000th are extracted. Yes. In step S14, as in step S4, the arrangement list L0 in which the thumbnails of the frame images G01 to G08 are arranged in a row is displayed as a list in the form shown in FIG. 5 or the arrangement of thumbnails in another arrangement shape. Has been.

  Subsequently, the process proceeds to step S15 (part of the narrowing-down process), and in a manner similar to step S5, a touch input is performed on the displayed array list L0, or a standby state waits for another input to be performed. enter. The user examines the array list L0 and performs input for selecting a portion including the start point and the end point of the target range (the range to be clipped / deleted). For example, in the present embodiment, by tracing on the touch panel 4 a range from a thumbnail that seems to be closest to the start point to a thumbnail that seems to be closest to the end point, it is possible to perform input for specifying this range.

  If touch input for selecting a moving image range is performed in the standby state in step S15, the control unit 7 proceeds to step S16. In step S16, the control unit 7 determines the selected portion based on the content of the touch input. For example, as shown by an arrow B in FIG. 6A, when an input that traces a section from the frame images G04 to G06 on the array list L0 is made, the control unit 7 determines the traced section based on this input. Extract each range near the start and end points. Specifically, a range (first range) defined by the beginning of the traced section and the immediately preceding thumbnail is extracted as a range around the start point. In the example of FIG. 6A, the range from the frame images G03 to G04 is extracted as the selection range. On the other hand, a range (second range) defined by the end of the traced section and the thumbnail immediately after that is extracted as a range around the end point. In the example of FIG. 6A, the range from the frame images G06 to G07 is extracted as the selection range. Alternatively, when the range of the moving image to be selected is narrow, the range can be specified by touching only one thumbnail. In this case, since there is only one thumbnail in the designated range, this thumbnail is handled as the first and last thumbnail. That is, the section between the designated thumbnail and the thumbnail immediately before it can be treated as a range including the start point, and the section between the designated thumbnail and the thumbnail immediately after it can be treated as a range including the end point.

  Subsequently, the control unit 7 proceeds to step S17 (part of the narrowing-down process), and performs a process of resetting the above two selection ranges as a new population. Then, the process proceeds to step S18. In step S18, as in step S2, if the number of frame images constituting the population is greater than N (step S18: Yes), the process proceeds to step S19. In step S12 and step S18, if the number of frame images in the population is N or less (step S12: No / step S18: No), there is no need to narrow down the frame images, so step that is the final process of the search. Proceeding to S24, the entire population is displayed in a list as in step S8.

  When the control unit 7 proceeds to step S19 (extraction process), N / 2 pieces are respectively obtained from the first range (near the start point) and the second range (near the end point) which are the new population set in step S17. Thus, a total of N frame images are extracted. In other words, since N = 8 in this embodiment, four frame images are extracted from each range. As a method for extracting frame images from each range, the same method as in step S3 is used. That is, the start point and end point of each range, and each range are divided into three (N / 2-1 division) at equal intervals, the positions of the two division points are calculated, and the frame image at each division point is extracted.

  FIG. 6B shows an example of the extraction result obtained in step S19. As described above, since the first range is the range from the frame image G03 (15428th) to G04 (23142th), the 15428th, 17999th, 20570th, and 23142th four frame images G11 to G14 are included. Extracted. Similarly, since the second range is a range from the frame image G06 (38571th) to G07 (46285th), the 38571th, 41142th, 43713th, and 46285th frame images G15 to G18 are extracted. The

  Next, the control unit 7 proceeds to step S20 (list display step), and generates thumbnails for each of the extracted N (eight in this embodiment) frame images in the same manner as steps S4 and S14. Are displayed in a list on the input display panel 2. Then, it progresses to step S21 (a part of narrowing-down process), and enters a standby state. And if the input for range selection is performed, it will progress to step S22, S23 based on the said input. In steps S22 and S23, processing similar to that in steps S16 and S17 is performed to further narrow the range around the start point and the end point based on the input result. Then, the process returns to step S18. Thereafter, similarly, until the number of frame images in the population is N or less (step S18: No), or other input than the range designation is performed in step S21 (standby state), S18 to S23. Repeat the process. As a result, the moving images near the start point and the end point are further narrowed down.

  When the list display process in step S20 is performed, the user confirms the thumbnail arrangement list Ln (n = 0, 1, 2,...) Displayed in this step. When it is determined that the thumbnails of the frame images to be set as the start point and the end point are included in the array list Ln, an input for determining them as the start point and the end point is input in the standby state in step S21. It can be carried out. When this input is performed, the control unit 7 proceeds to step S26, and determines the start and end frame images based on the input content. Then, it is determined that the search is finished, and the processing of this flowchart is finished. Then, the search result is confirmed, or the clipping process or the deletion process is performed based on the search result. Similar to step S5, other input can be performed to interrupt or change the search process as appropriate.

  When the processing of steps S18 to S23 is repeated without determining the start point and the end point, each time the number of frame images included in the population follows that time, the magnification is 1 / ((N / 2) -1). It decreases exponentially and the frame image extraction interval narrows exponentially. FIG. 6C shows an example of the third extraction result. With the selection of a range from G13 (20570th) to G16 (41142th), frame images G21 to G28 are extracted. As shown in FIG. 3, in the first extraction process, the extraction interval of the frame images G01 to G08 is every 7714 frames, and every second is 2571 frames, whereas every third time is every 857 frames. If the processing is continued in the same way, every 285 frames for the fourth time, every 95 frames for the fifth time, every 31 frames for the sixth time, every 10 frames for the seventh time, every three frames for the eighth time, When a touch input for selecting a range is performed on the displayed thumbnail arrangement, the population is narrowed down to eight frame images. Therefore, the determination in the next step S18 is No, and the process proceeds to step S24, which is the final process of the search. That is, in step S24, the control unit 7 displays a list of all the populations in the same manner as in step S8, and then proceeds to step S25 to enter a standby state. Then, the process proceeds to step S26 based on the user input, and as described above, the process of determining the start point and the end point is performed and the search is terminated.

  As described above, in this search method, the number of frame images is narrowed down exponentially for each range around the start point and around the end point. Therefore, if the video is 30 minutes (the number of frame images is 54000), It is possible to proceed to the final stage of the frame image search only by performing selection input for display eight times at the maximum (when N = 8). Therefore, as with the division position search of (1), the start point and the end point can be detected easily and efficiently with a much smaller amount of operation than in the prior art, and the processing load is reduced. In addition, this method is more efficient because the start point and end point can be searched simultaneously.

  Note that this search method is a method of searching for the start point and end point of one range, but the search target may be two unrelated frame images other than the start point and end point. For example, instead of performing range input on the thumbnail arrangement list, two-point touch input is performed, and two frame images can be searched simultaneously. In addition, the above method extracts a range of videos defined by two adjacent thumbnails at the beginning and immediately before the traced range, and at the end and immediately after the traced range, and sets it as a new population. However, a range defined by two sets of adjacent thumbnails at the beginning and immediately after the traced range, and at the end and immediately before the traced range may be extracted. Moreover, you may extract the moving image part of another range based on the traced range.

(Embodiment 2)
In the second embodiment, the search method of the present invention is applied to an adjustment value determination method (adjustment value search method) in image processing. This method can be executed by storing an image processing application program that provides a function of executing the search method of the present embodiment in an apparatus having the same configuration as the moving image editing apparatus 1 described above. This program is read into the control unit, input for selecting an image to be edited from the list of image data is performed, and image data for image processing is read.

  The image processing application program assumes a large number of candidate images in which the value of the feature value is changed little by little when performing the process of changing the feature value in the image. The target image processing is efficiently performed by searching for the image. As the feature amount, various values such as an index value indicating the hue, brightness, saturation (contrast), resolution, or the degree of other effects of the image are assumed.

(Adjustment value determination method for image adjustment (Adjustment value search method))
First, a population is set as in step S1 of the first embodiment. The image processing application sets the entire adjustable range of the feature amount to be adjusted or a part thereof as a population. Here, when the feature amount is adjusted discretely, the number of candidate images (the number of images constituting the population) is limited, so the number of processes until the final selection is made according to the number of images constituting the population. Is decided. Therefore, in this case, similar to the division position search method in the first embodiment, extraction of candidate images from the population (extraction process), display of an array list of thumbnails (list display process), range selection, and population The range of the feature value can be narrowed down until the target image appears in the array list. On the other hand, if the feature amount can be changed continuously, the number of candidate images is theoretically infinite. Therefore, in this case, the determination of the number of constituent elements of the population performed in step S2 of the first embodiment is omitted.

  FIG. 8 shows a display example of an array list when hue adjustment is performed. FIG. 8A is a display example of the sequence list extracted by the first extraction process. In this way, in the first list display step, an image whose hue is changed little by little is extracted and displayed in a ring shape along the hue circle. In the case of hue adjustment, the position on the hue circle can be indicated by an angle value. Therefore, in the first extraction, the value of each division point obtained by equally dividing the 360 degree angle range by the number of thumbnails that can be displayed is calculated. A thumbnail image of a hue corresponding to each division point is generated and displayed in an array. FIG. 8B is a display example of the sequence list extracted by the second extraction process. When the position indicated by the arrow C1 in FIG. 8A is touch-inputted, a predetermined angle range including this position is set as a population, and the extracted thumbnail is displayed in an arc shape in the second list display step. can do. In addition, as shown by the arrow C2 in FIG. 8A, a predetermined angle range may be input and this angle range may be set as the next population. Moreover, the arrangement | sequence shape at the time of displaying can be changed suitably, and can also be made into linear form and another shape.

  When a thumbnail of a desired image is displayed, the user can select the thumbnail to end or interrupt the search. The search may be interrupted, the image may be confirmed on a large screen or the like, and it may be possible to select whether to perform a further search or to end the search here.

  As described above, in the method according to the second embodiment, when adjusting the feature amount of an image, a list of images in which the feature amount is changed little by little can be displayed. Therefore, the adjustment value can be determined more efficiently than in the conventional adjustment method in which the adjustment of the feature amount and the image confirmation are alternately performed many times and trial and error. In addition, since the adjustment value can be determined without performing fine and labor-intensive operations such as fine adjustment by scrolling and input of the adjustment value, image adjustment can be performed easily and efficiently. Note that the function of the image adjustment application may be incorporated in the moving image editing application of the first embodiment, and the image adjustment for the frame image being edited may be performed by this method.

(Embodiment 3)
In the third embodiment, the search method of the present invention is used to search for operating flights of transportation facilities. This method can also be executed by storing a flight search application program that provides a function of executing the search method of the present embodiment in an apparatus having the same configuration as the moving image editing apparatus 1 described above. By reading this program into the control unit and inputting the route, boarding section, etc., the corresponding data is read from the timetable data stored in advance. The flight search application program searches the target flight from the read flight flight list by the search method described below.

(Operating flight search method)
First, a population is set as in step S1 of the first embodiment. When a boarding date, a rough boarding time range, etc. are input in addition to the above-mentioned route and boarding section, the flight search application program sets all data of the corresponding flight to the population. The flight search application program extracts the flight candidate from the population (extraction process) and arranges the flights in time series for this population, in the same way as the division position search method in the first embodiment. Sequence list display (list display process), range selection and population resetting (narrowing process) are performed. Then, it is possible to narrow down the service flights until the target service flight appears in the array list.

  When a flight search is performed, an array list in which displays showing information on flight flights are arranged in time series is generated and displayed. The operation flights to be displayed can be extracted as a flight closest to each dividing point obtained by equally dividing the departure / arrival time range of the population and each flight having the earliest time and the latest time in the population. The arrangement direction can be various directions such as one horizontal row, one vertical row, and one diagonal row. When a flight that meets the desired condition is displayed, the user can select the flight and terminate or stop the search. The search may be interrupted to display detailed information on the selected flight, and it may be possible to select whether to perform a further search or to end the search here.

  As described above, in the method according to the third embodiment, when a large number of flights are set, it is possible to efficiently perform the work of narrowing down from these. In addition, when searching for arrival flights at departure positions and departure flights from them in connection search, etc., if the service flight search of this embodiment can be used as an auxiliary, it can be efficiently requested. You can find the right combination of flights.

  The field of use of the search method of the present invention is not limited to Embodiments 1 to 3 above, and can be widely applied to a search for a series of candidate data that can be arranged according to a predetermined arrangement standard. That is, a process for setting a series of candidate data that can be arranged according to a predetermined arrangement standard in a population, a process for extracting a preset number of candidate data according to an extraction rule set in advance, and extraction results as thumbnails, etc. The process of displaying the list side by side in the form of the representative display is performed to enter a standby state for allowing the user to select a part of the displayed array display. When a selection input is made, the population is reset so that the portion of the population corresponding to the selected portion becomes a new population. After that, the search is terminated or interrupted because the target display is found in the representative display displayed in the list, or the number of set populations is small. Until it is determined that a further extraction process is not necessary for the reason, a process of extracting a preset number of candidates, a list display process, and a population narrowing process are repeated. By this search method, target candidate data can be efficiently searched from various candidate data.

(Modification example)
(1) In each of the above embodiments, when extracting N candidate data (frame image, image of adjustment process, service flight, etc.) for list display from the population, the range of the population is equal to N-1. The candidate data corresponding to each dividing point is extracted and the candidate data is added to the candidate at the beginning and end of the population. However, candidate data can be extracted according to other extraction rules. . For example, a division ratio different from the equal division may be set in advance, the population may be divided according to the division ratio, candidate data for each division point may be extracted, and candidate data including these may be displayed as a list. For example, at various positions such as the vicinity of the beginning and the end of the population, the width of the divided section can be made narrower or wider than other positions. It is also possible to add rules such as extracting specific candidate data with priority or excluding them. For example, in the example of the third embodiment, when there are an express flight and a normal flight immediately before and after the division position, a rule may be provided such that the express flight is preferentially extracted.

(2) In each of the above embodiments, candidate data is selected by input to the touch panel 4, but various input methods such as moving the cursor displayed on the screen using a mouse, touchpad, keyboard, or the like. Selection input can be performed.

(3) In each of the above embodiments, the search method according to the present invention is used to advance to the final stage of the search. Moreover, usability can be made favorable by making it possible to return to the previous process as appropriate and increasing the degree of freedom.

DESCRIPTION OF SYMBOLS 1 ... Moving image editing apparatus 2 ... Input display panel 3 ... Liquid crystal display 4 ... Touch panel 5 ... Case 6 ... CPU
7. Control unit 8 ... Storage unit 9 ... Video controllers F01-F08 ... Frame images F11-F18 ... Frame images F21-F28 ... Frame images G01-G08 ... Frame images G11-G18 ... Frame images G21-G28 ... Frame images L0, L1, L2 ... Sequence list R0, R1, R2 ... Sequence list

Claims (14)

Based on an input for designating a search target or instructing start of search, the control means performs a process of setting a series of candidate data that can be arranged according to a predetermined sequence standard as a population to be searched; and
An extraction step in which the control means performs a process of extracting a preset number of candidate data according to a preset extraction rule from the population;
A list display step in which the control means generates a sequence display in which the display indicating the extracted candidate data is arranged in a line in accordance with the sequence criteria, and performs a process of displaying the sequence display on the display means as a list;
The control unit waits for a selection input for selecting a part of the array display, and when the selection input is performed, which part of the array display is selected based on the selection input. And performing a process of setting a portion of the population corresponding to the selected portion as a new population,
The control means, when the population is set, until an input for ending or interrupting the search is performed, or until it is determined that the population does not need to perform the extraction step A search method characterized by repeating the three steps of the extraction step, the list display step, and the narrowing step. In claim 1,
When the control unit determines that the population includes only discrete candidate data equal to or less than the preset number, the control unit stops the repetition of the three steps, and all candidates constituting the population A search method characterized by performing a process of arranging a display showing data in a line and displaying the list on the display means. In claim 1 or 2,
The control means corresponds to each division position when an array in which all candidate data constituting the population is arranged in a line according to the arrangement standard is divided into a predetermined number at a preset division ratio in the extraction step. A search method characterized by extracting the preset number of candidate data so as to include candidate data. In claim 1 or 2,
The control means, in the extraction step, as the preset number of candidate data, the candidate data at the beginning and the end of the sequence in which all candidate data constituting the population are arranged in a line according to the sequence criterion, and A search method characterized by extracting candidate data corresponding to each division position when the array is equally divided into a predetermined number. In any one of claims 1 to 4,
The control means, in the list display step, the arrangement direction of the arrangement display to be displayed in the list is arranged in a straight line in one of vertical, horizontal, and diagonal directions, or arranged in an arc or ring A search method characterized by an array direction. In any one of claims 1 to 5,
In the case where the series of candidate data is a moving image composed of a number of frame images that are sequentially switched and displayed every predetermined time,
The control means includes
In the extraction step, a process of extracting the preset number of frame images from the preset number of extraction positions set to be distributed over the entire range of the moving image set as the population. Done
In the list display step, a process of generating an array display in which thumbnails of the frame images extracted in the extraction step are arranged in a line according to the display order of the frame images, and displaying the array display on a single screen as a list Done
In the narrowing-down process, when it is determined that a part of the array display is selected, a process of setting the range of the moving image corresponding to the selected part to a new population is performed,
When the population is set, input for ending or interrupting the search is performed, or until it is determined that the number of frame images constituting the population is equal to or less than the preset number. Repeat the three steps of the extraction step, the list display step, and the narrowing step,
When the control means determines that the number of frame images constituting the population is equal to or less than the preset number, the control means displays a list of array displays in which thumbnails of all the frame images constituting the population are arranged in a line. A search method characterized by performing processing. In claim 6,
In the narrowing-down process, when the control unit determines that a portion between two adjacent thumbnails in the array display is selected, one of the two frame images corresponding to the two adjacent thumbnails is set as a starting point, A search method, characterized in that the range of the moving image having the other end point is set as the new population. In claim 6,
The control means includes
In the narrowing step,
When it is determined that a section from any thumbnail to another thumbnail in the array display is selected,
The two frame images corresponding to the thumbnail located at the head of the selected section and the thumbnail located immediately before it are set as the start point and end point of the first range in the moving image, and at the end of the selected section Two frame images corresponding to the thumbnail located at and the thumbnail located immediately thereafter are set as the start point and end point of the second range in the moving image, and the first and second ranges are set as the new range. Set as a population,
In the extraction step immediately after setting the first and second ranges as a new population,
The preset number of the frame images are extracted from the preset number of extraction positions set so as to be distributed by a predetermined number within each of the first and second ranges. retrieval method. In any one of claims 1 to 5,
In the case where the series of candidate data is data of a flight group of transportation facilities,
The control means includes
In the extracting step, the predetermined number of service flights is included so as to include data of each service flight closest to each division time point obtained by dividing the departure / arrival time range of the service flights into a predetermined number at a predetermined division ratio. A search method characterized by extracting data. In any one of claims 1 to 5,
When the series of candidate data is data of an image group corresponding to each value of the feature amount when a predetermined feature amount in the image is changed within a predetermined range,
The control means includes
In the extracting step, the predetermined number of data is included so as to include data of each image corresponding to each division point when the change range of the feature amount in the image group is divided into a predetermined number at a predetermined division ratio. A search method characterized by extracting image data. A search device for searching for target candidate data from a population to be searched by the search method according to any one of claims 1 to 10,
Display means for displaying a list of array displays in which a predetermined number of displays are arranged in a line;
Input means for receiving a selection input for selecting a part of the array display;
Storage means for storing a program for executing the search method;
And a control unit that executes processing for performing each step in the search method based on a program read from the storage unit. In claim 11,
The display means is a liquid crystal display panel;
The input means is a touch panel installed on the surface of the liquid crystal display panel,
The control means includes
In the narrowing-down process, when an operation of touching or tracing a part of the array display displayed as a list on the liquid crystal display panel is performed, the touched part or the part of the population corresponding to the traced part Is set as a new population. A moving image editing apparatus for searching for a frame image of an edited portion in a moving image by the search method according to any one of claims 6 to 8,
Display means for displaying a list of array displays in which a predetermined number of thumbnails of frame images included in the video data are arranged in a line on a single screen;
Input means for receiving a selection input for selecting a part of the array display;
Storage means for storing a program for executing the search method;
And a control unit that executes processing for performing each step in the search method based on a program read from the storage unit. In claim 13,
The display means is a liquid crystal display panel;
The input means is a touch panel installed on the surface of the liquid crystal display panel,
The control means includes
In the narrowing-down process, when an operation of touching or tracing a part of the array display displayed as a list on the liquid crystal display panel is performed, the touched part or the part of the moving image data corresponding to the traced part is displayed. A moving image editing apparatus characterized by being set as a new population.

Images