JP2009141882A - Imaging apparatus and its image automatic selection method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To automatically extract and select a scene a user pays attention to in moving image data during a period of photographing the moving image. <P>SOLUTION: In an image automatic selection method of an imaging apparatus that stores moving image data obtained by photographing a subject in a recording medium by compressing it to a reference frame or a differential frame for each frame according to a moving image compression standard while photographing the subject, feature values between a reference frame and a differential frame continuous with each other is detected and the reference frame or the differential frame whose amount of features is greater than or equal to a threshold is selected after adding selection information thereto. As the feature values, a value indicative of a correlation of data size between a reference frame and a differential frame is used, and preferably, the feature values is used, which becomes a greater value as the difference in data size or its ratio is smaller. With this arrangement, it is made possible to extract scenes of the subject at "end of movement" and "start of movement". <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to an image pickup apparatus such as a digital camera, and more particularly to an image pickup apparatus capable of automatically selecting and displaying individual images intended by a user from a series of images picked up by a continuous shooting function or a moving image shooting function. And an image automatic selection method thereof.

  When shooting video images with a video camera or the like, some have the function of selecting and playing back several images in the video image in order to quickly and easily check what video was shot. is there.

  For example, in the prior art described in Patent Document 1, a priority that automatically represents the importance of an image is automatically determined based on a camera button operation performed at the time of shooting, and an image to be displayed is selected according to the priority. The digest is displayed in accordance with the user's intention at the time of shooting.

  In the prior art described in Patent Document 2, two frames of frame memory are prepared, and two consecutive frame images stored in each frame memory are compared to calculate still image data when the scene changes. This is automatically recorded together with the address information. Further, in the prior art described in Patent Document 3, a frame having a large difference in luminance value between adjacent frames is extracted from a series of moving image data of a plurality of frames, and this frame image is used as a representative image.

JP-A-6-165209 JP-A-6-217254 JP-A-9-233422

  The prior art described in Patent Document 1 requires a user to operate a camera button when selecting an image, and if the user forgets to operate the button, an appropriate image cannot be selected.

  The prior art described in Patent Document 2 requires a comparison operation between frames, so that there is a problem that the calculation processing load becomes large and the memory for the comparison operation becomes large and costs increase.

  The prior art described in Patent Document 3 also has a problem that the calculation load increases because the luminance difference between adjacent frames is calculated.

  An object of the present invention is to provide an imaging apparatus capable of automatically selecting and displaying an appropriate image without a user's special button operation and with a small processing load, and an image automatic selection method thereof. There is.

  In the image automatic selection method of the imaging apparatus of the present invention, an image of the imaging apparatus that compresses moving image data obtained by capturing a subject into a reference frame or a difference frame for each frame according to a moving image compression standard and stores the same in a recording medium. In the automatic selection method, a feature amount between the continuous reference frame and the difference frame is detected, and selection information is added to the reference frame or the difference frame in which the feature amount is equal to or greater than a threshold value.

  The image automatic selection method for an imaging apparatus according to the present invention is characterized in that a still image of a frame to which the selection information is added is displayed on a display unit.

  The feature amount in the image automatic selection method of the imaging apparatus according to the present invention is a value representing a correlation in data size between the reference frame and the difference frame.

  The feature amount in the image automatic selection method of the imaging apparatus according to the present invention is such that the smaller the difference or ratio of the data sizes, the larger the feature amount.

  The image automatic selection method of the imaging apparatus according to the present invention is characterized in that when there are a plurality of types of frames as the difference frame, the feature amount is obtained by weighting each frame type.

  The image automatic selection method of the imaging apparatus according to the present invention is characterized in that a priority is added to the selection information in order of the feature amount.

  The image automatic selection method of the imaging apparatus according to the present invention is characterized in that the number of frames to which the selection information is added is limited to a predetermined number in the order of priority.

  In the image automatic selection method of the imaging apparatus according to the present invention, when the compressed data of the still image is generated in addition to the compression of the moving image data, the still image stored in the recording medium is determined based on the feature amount and the selection is performed. Only a still image of a frame to which information is added is stored.

  In the image automatic selection method of the imaging apparatus of the present invention, before detecting a feature amount between the reference frame and the difference frame for each frame, the feature between consecutive frame groups in units of a predetermined number of frame groups. An amount is obtained, a frame group in which the feature amount is equal to or greater than a threshold value is selected, and a feature amount between the reference frame and the difference frame is detected for each frame of the selected frame group.

  The feature amount between the frame groups in the image automatic selection method of the imaging apparatus according to the present invention is obtained between the reference frames included in each frame group.

  In the image automatic selection method for an image pickup apparatus according to the present invention, the feature amount between the frame groups is a feature amount obtained from the whole of each frame group.

  The image pickup apparatus of the present invention is an image pickup apparatus that compresses moving image data obtained by shooting a subject into a reference frame or a difference frame for each frame according to a moving image compression standard and stores the data in a recording medium. Control means for executing the described automatic image selection method is provided.

  According to the present invention, the scene in which the user pays attention in the moving image data is automatically extracted and selected during the shooting of the moving image, and the processing load when reproducing the still image of the scene is reduced and at the time of reproduction. The waiting time becomes shorter and smooth playback becomes possible.

  Hereinafter, an embodiment of the present invention will be described with reference to the drawings.

  FIG. 1 is a functional block diagram of an imaging apparatus according to an embodiment of the present invention. An imaging device 1 having a moving image shooting function includes a shooting lens 2 that forms an image of the object scene light, and an optical image of the object scene light that is provided on the back of the shooting lens 2 and is converted into an electrical signal by the light receiving surface and output. Output from the solid-state imaging device 3, an analog front end (AFE) circuit 4 that performs correlated double sampling processing, gain control, and analog / digital conversion processing on the output signal (captured image signal) of the solid-state imaging device 3, and output from the AFE circuit 4 And a digital signal processor (DSP: control means for selecting an image) 5 that takes in a digital captured image signal and performs various image processing.

  A dedicated video compression circuit 6 is connected to the DSP 5, and when the DSP 5 transmits the moving image data captured from the AFE circuit 4 and the audio signal captured from a microphone (not shown) to the dedicated video compression circuit 6, the dedicated video compression circuit 6 The received moving image data and audio signal are converted into MPEG2, MPEG4, H.264. Data compression is performed according to a predetermined moving image compression standard such as H.264, and the compressed data is returned to the DSP 5.

  A memory 7 used by the moving picture compression dedicated circuit 6 is connected to the moving picture compression dedicated circuit 6, and a detachable external recording medium 8 for recording and storing the compressed captured image data and audio data is connected to the DSP 5. Is done.

  A data bus 9 is connected to the DSP 5, an SDRAM 10 used as a main memory by the DSP 5 and a flash ROM 11 storing programs and the like are connected to the data bus 9, and a liquid crystal monitor screen provided on the back surface of the imaging device 12 is also connected to the data bus 9.

  FIG. 2 is a diagram illustrating an example of a moving image. The illustrated example shows a series of images where a rabbit runs, stops, and starts running. Considering what the user is interested in in this series of images, it is the images 15 and 16 of the “end of movement” and “start of movement” of the rabbit shown by the thick black frames 15 and 16. There are often.

  Therefore, in the present embodiment, from among a series of moving image data, the image of “end of movement” and “start of movement” is automatically selected during shooting and selection information is added, The image with the selection information is set as a reproduction target image when the digest version of the moving image is reproduced.

  FIG. 3 is an explanatory diagram of encoded data generated by the moving image compression standard. The encoded data is configured in GOP units as shown in the figure. The GOP unit refers to a frame group composed of a series of difference frames based on one I frame, where the I frame is a reference image, the P frame is a difference image from the nearest past I frame or P frame, The B frame is a difference image from the nearest past and future I frames or P frames.

  FIG. 4 shows an I frame in each compressed data (encoded data) when the amount of angle of view difference between two consecutive frames is 20 pixels (small motion) and 200 pixels (large motion). , B is a graph showing the code amount of B frame.

  According to this graph, it can be seen that if the motion between two consecutive frames is small (20 pixels), the difference in data size between the two frames is large. When the movement is small, the data size of the difference image of the B frame becomes small with respect to the data size of the reference image of the I frame, so that the data size difference becomes large.

  Conversely, when the motion between two consecutive frames increases, the data amount of the difference image of the B frame indicating the changed portion with respect to the reference image of the I frame increases, and the difference in data size between the two frames becomes Get smaller.

  As described with reference to FIG. 2, an image showing a change of “end of movement” or “start of movement” is an image having a large movement with respect to the preceding and succeeding images. Can be identified.

  That is, in this embodiment, the magnitude of motion is determined from the correlation of the data size between two frames, and an image to be reproduced is selected. That is, from the correlation that the smaller the motion, the larger the difference in data size between the I frame and the B frame, the size of the motion is determined, and an image with a large motion is automatically selected as a digest version playback image.

  Hereinafter, a feature amount expressed as a function of the absolute value of the difference in data size between two frames is referred to as a “score”, and the score increases as the difference in data size decreases. Alternatively, a feature amount represented by a function of a data size ratio between two frames is referred to as a “score”, and the score increases as the ratio decreases.

  FIG. 5 is a flowchart showing a processing procedure of a program for executing the image automatic selection method according to the first embodiment of the present invention. This program is read from the ROM 11 when the DSP 5 processes the moving image data received from the moving image compression dedicated circuit 6, and is executed in parallel with the processing of the moving image data.

  This program is activated when the DSP 5 passes image data to the moving image compression circuit 6, and the moving image compression circuit 6 encodes the received image data in accordance with a predetermined moving image compression standard (step S1). .

  The moving image compression dedicated circuit 6 that performs encoding (moving image compression) transmits the compressed moving image data (encoded data) to the DSP 5. At this time, the DSP 5 determines whether or not one packet of compressed data has been received from the moving image compression dedicated circuit 6 (step S2). If reception of one packet has not been completed, step S2 is repeatedly executed.

  If one packet of compressed data is received, it is next determined whether or not the one packet data includes a flag indicating the head of the frame (step S3). If it is not the head of the frame, the process returns to step S2. When it is the head of the frame, it can be seen that the packet group received up to the previous frame indicates the previous frame.

  Therefore, if it is the head of the frame, the process proceeds from step S3 to step S4, the data sizes of the packet groups of the previous frame and the current frame are compared, and a score is calculated. Then, in the next step S5, it is determined whether or not the score value is greater than or equal to a predetermined threshold value.

  If the score value is not equal to or greater than the predetermined threshold, it can be determined that the motion between the two frames of the image is small as described with reference to FIG. 3, and the change image is not “end of motion” or “start of motion”. Return.

  If the score value is equal to or greater than the predetermined threshold value, the process proceeds to step S6, where the image of the previous frame can be determined as the “end of movement” image or the “start of movement” image. Select as an image.

  This selection is performed, for example, by attaching a selection flag to this image, or by writing a frame number indicating the image in a predetermined area of the recording medium, or writing it as header information of moving image data.

  After the digest version playback target image is selected in step S6, in the next step S7, it is determined whether or not the video compression encoding has been completed. If not, the process returns to step S2 and is completed. In this case, the process of FIG.

  As described above, according to the present embodiment, since it is selected whether or not to be a digest version playback target image during shooting, the processing load during actual playback is reduced, and there is no waiting time and smoothness. Playback is possible.

  Note that the score value can be calculated based on an index expressing the magnitude of motion, such as the data size itself, the difference in data size between adjacent frames, and the like. Differences in data size due to differences in frame types such as I, P, and B can be absorbed by performing processing such as multiplication by a predetermined coefficient.

  FIG. 6 is a flowchart showing a processing procedure of a program for executing the image automatic selection method according to the second embodiment of the present invention. The processing procedure of this embodiment is the same up to step S4 of the procedure shown in FIG. 5, and when the score between two frames is calculated, this score value is used as the score value of the previous frame of the two frames. Corresponding to the frame, it is temporarily stored in the main memory 10 or the like.

  Then, in the next step S9, the score values temporarily stored are sorted, and in the next step S10, it is determined whether or not the encoding is completed. Returning to S2, and if completed, proceed to step S11.

  In step S11, the priority order data temporarily stored in the main memory 10 is written in the recording medium 8 or written as header information of the moving image data, and then this process is terminated.

  In this way, in this embodiment, the score value of each frame is held as information sorted during shooting. Therefore, if the user designates the number of images to be reproduced at the time of reproduction, the number of reproduced images corresponding to the number of images is waited according to the sort order. Playback and display can be performed smoothly without time.

  In addition, according to the present embodiment, when the number of images to be reproduced is preset to a default value or a predetermined number specified by the user, for example, 20 images, the process of step S9 can be performed up to the top 20 ranks. .

  FIG. 7 is a flowchart showing a processing procedure of a program for executing the image automatic selection method according to the third embodiment of the present invention.

  Some digital cameras that capture moving images can automatically capture still images in the middle of video recording. In this case, instead of shooting a still image separately from the moving image, the captured image data to be a still image is extracted from the moving image data before compression, and compressed according to a standard such as JPEG format separately from the moving image data. become.

  In this case, the digest version playback image of the moving image may be selected from the generated still image data. However, all the generated still image data together with the moving image data is stored in a buffer memory or a recording medium in terms of memory capacity. Sometimes it cannot be saved. Therefore, still image data to be stored is selected from the generated still image data. For this selection, a score value obtained from compressed moving image data is used as described above.

  First, the minimum score counter Smin is cleared to 0 (step S21), the reproduction number counter NUM is cleared to 0 (step S22), and moving image data encoder (compression) is started (step S23).

  Next, the DSP 5 determines whether or not data for one packet has been received (step S24), and determines whether or not a frame head flag is included in the data for one packet (step S25). If the head flag is included, a score value S for frame selection is calculated (step S26). Up to this point, the processing is the same as in the above-described embodiment.

  Next, when the score value S is the score value of the first frame, the score value S is set to Smin (step S27), and it is determined whether or not the score value S is equal to or greater than a threshold value (step S28). If the score value S is greater than or equal to the threshold value, still image data corresponding to the frame is set as a reproduction target (step S29). If the score value S is not greater than or equal to the threshold value, the process returns to step S24.

  Following step S29, the reproduction number counter NUM is counted up at step S30, and at the next step S31, it is determined whether or not the score value S is equal to or greater than Smin.

  If the score value S is less than Smin, the process proceeds from step S31 to step S32, where the value of Smin is replaced with a score value S smaller than that, and in the next step S33, the score value S of the previous frames is the smallest. The frame that becomes is the previous frame, and the process proceeds to step S34. Moreover, also when score value S is more than Smin as a result of determination of step S31, it progresses to step S34.

  In step S34, it is determined whether or not the count number of the counter NUM is equal to or smaller than a certain value, that is, whether or not the number of reproductions is equal to or less than a predetermined number. If the predetermined number has been exceeded, the process proceeds to step S35, the frame Fmin having the minimum score value S is excluded from the reproduction target, and the reproduction number counter NUM is counted down in the next step S36, and the process returns to step S24.

  As a result of the determination in step S34, if the count number of the counter NUM is equal to or smaller than a predetermined value, the process proceeds to the next step S37, and it is determined whether or not the number of moving image frames to be processed is a multiple of a predetermined value. . In moving image shooting, since a predetermined number of shots are taken at fixed time intervals, the process returns from step S37 to step S21 when a multiple of the predetermined value is reached, that is, every time the fixed time period elapses. As a result, it is possible to limit the number of images to be reproduced to a predetermined number for reproduction every certain time.

  If the result of the determination in step S37 is not a multiple of the predetermined value, the process proceeds to step S38, where it is determined whether or not the encoding has been completed. If it has not been completed, the process returns to step S24. This process ends.

  As described above, according to the present embodiment, when the number of still images exceeds the predetermined limit number, the reproduced image having the smallest score at that time is set to non-reproduced, and therefore within a certain time. Thus, it is possible to control so that the buffer for storing still image data does not overflow. The limit number is determined from the access speed of the recording medium, the compression rate of still image data, and the like.

  FIG. 8 is a flowchart showing a processing procedure of a program for executing the image automatic selection method according to the fourth embodiment of the present invention. In the embodiment described above, it is determined whether or not the image to be reproduced is set for each frame, but in this embodiment, the selection is made in two stages. First, it is determined whether or not to be a playback target for each GOP, and then a playback target image is selected from the frame group of the GOP determined to be the playback target.

  When the encoding is started (step S41), it is next determined whether or not the encoding of the I frame is completed (step S42). When the encoding of the I frame is completed, the data size of the I frame is read next (step S43), and a function of the difference between the I frame size of the previous GOP and the I frame size of the current GOP is used as a feature amount. A score is set (step S44).

  In the next step S45, GOP selection determination processing is performed. If it can be determined that the GOP has a large score value and a change in motion, the current GOP is selected. If it can be determined that the GOP has a small score value and a small change in motion, the current GOP is not selected.

  In the next step S46, it is determined whether or not the current GOP is selected. If not selected, the process returns to step S42. If selected, the process proceeds to step S47, and a frame indicating “end of movement” and “start of movement” is processed in the same group as the embodiment described with reference to FIGS. Select by. Then, it is determined whether or not the encoding is completed (step S48). If it is not completed, the process returns to step S42, and if it is completed, this process is terminated.

  This embodiment is premised on image compression by CBR (Constant Bit Rate System). In this case, in a scene with a large motion, a large amount of data is allocated to a differential image such as a B frame or a P frame, and the I frame size is small. Become. A score value indicating the magnitude of the motion is calculated using the property.

  According to the present embodiment, it is possible to increase the selection efficiency by including a step of roughly selecting moving image data captured by the fixed bit rate method.

  FIG. 9 is a flowchart showing a processing procedure of a program for executing the image automatic selection method according to the fifth embodiment of the present invention. In the fourth embodiment, the determination process for each GOP is performed on the I frame image. However, the present embodiment is different in that the determination is performed based on the data size of the entire GOP.

  When encoding is started (step S51), it is next determined whether or not encoding of 1 GOP unit is completed (step S52). When encoding of 1 GOP unit is completed, the data size of the entire GOP is read next (step S53), and a function of the difference between the overall size of the previous GOP and the current GOP is used as a feature amount, and this is used as a “score”. (Step S54).

  In the next step S55, GOP selection determination processing is performed. If it can be determined that the GOP has a large score value and a change in motion, the current GOP is selected. If it can be determined that the GOP has a small score value and a small change in motion, the current GOP is not selected.

  In the next step S56, it is determined whether or not the current GOP has been selected. If not selected, the process returns to step S52. If selected, the process proceeds to step S57, and a frame indicating “end of movement” or “start of movement” is selected from the group of frames of the current GOP, as in the embodiment described with reference to FIGS. Select by. Then, it is determined whether or not the encoding is completed (step S58). If the encoding is not completed, the process returns to step S52. If the encoding is completed, the process is terminated.

  This embodiment is based on the premise of video compression using VBR (variable bit rate method), and in this case, a large code amount is given to a scene with a large motion. Utilizing this property, the data size in units of GOP is used as an index indicating the magnitude of motion.

  According to this embodiment, in the variable bit rate method, the selection efficiency can be improved by including a step of roughly selecting.

  The feature amount “score” in GOP units described with reference to FIGS. 8 and 9 is determined according to the function of the appropriate data size difference depending on the moving image compression method.

  The automatic image selection method according to the present invention can automatically select a digest version playback image intended by the user from the moving image data, so that the usability of the digital camera equipped with the moving image shooting function is improved.

It is a functional block diagram of the digital camera which concerns on one Embodiment of this invention. It is a figure which shows an example of a moving image. It is a figure which shows the frame kind by a moving image compression standard. It is a graph which shows the difference of the data size of I frame and B frame in the case where a motion is large and small. It is a flowchart which shows the process sequence of the image automatic selection method which concerns on 1st Embodiment of this invention. It is a flowchart which shows the process sequence of the image automatic selection method which concerns on 2nd Embodiment of this invention. It is a flowchart which shows the process sequence of the image automatic selection method which concerns on 3rd Embodiment of this invention. It is a flowchart which shows the process sequence of the image automatic selection method which concerns on 4th Embodiment of this invention. It is a flowchart which shows the process sequence of the image automatic selection method which concerns on 5th Embodiment of this invention.

Explanation of symbols

3 Image sensor 5 Digital signal processor (DSP)
6 Video compression circuit 8 Recording media

Claims (12)

  In the image automatic selection method of an imaging apparatus in which moving image data obtained by photographing a subject is compressed into a reference frame or a difference frame for each frame according to a moving image compression standard during photographing and stored in a recording medium, the continuous reference frame and the continuous frame An image automatic selection method for an imaging apparatus, wherein a feature amount between a difference frame is detected, and selection information is added to the reference frame or the difference frame where the feature amount is equal to or greater than a threshold value.   2. The image automatic selection method for an imaging apparatus according to claim 1, wherein a still image of the frame to which the selection information is added is displayed on a display unit.   3. The image automatic selection method of the imaging apparatus according to claim 1, wherein the feature amount is a value representing a correlation of a data size between the reference frame and the difference frame.   4. The image automatic selection method for an image pickup apparatus according to claim 3, wherein the feature amount has a larger value as the difference or ratio of the data sizes is smaller.   5. The image automatic selection method for an imaging apparatus according to claim 1, wherein when there are a plurality of types of frames as the difference frames, the feature amount is obtained by weighting each frame type. .   6. The image automatic selection method for an imaging apparatus according to claim 1, wherein a priority order is added to the selection information in the order of the feature amount.   7. The image automatic selection method for an imaging apparatus according to claim 6, wherein the number of frames to which the selection information is added is limited to a predetermined number in the order of priority.   When generating still image compression data in addition to the compression of the moving image data, the still image to be stored in the recording medium is determined based on the feature amount, and only the still image of the frame to which the selection information is added is stored. The image automatic selection method for an imaging apparatus according to any one of claims 1 to 7,   Before detecting a feature amount between the reference frame and the difference frame for each frame, a feature amount between successive frame groups is obtained with a predetermined number of frame groups as a unit, and the feature amount is equal to or greater than a threshold value. The frame group is selected, and a feature amount between the reference frame and the difference frame is detected for each frame of the selected frame group. An image automatic selection method for an imaging apparatus.   The image automatic selection method for an imaging apparatus according to claim 9, wherein the feature amount between the frame groups is obtained between the reference frames included in each frame group.   The image automatic selection method for an imaging apparatus according to claim 9, wherein the feature amount between the frame groups is a feature amount obtained from the entirety of each frame group.   12. The imaging apparatus according to claim 1, wherein moving image data obtained by shooting a subject is compressed into a reference frame or a difference frame for each frame according to a moving image compression standard and stored in a recording medium during shooting. An image pickup apparatus comprising control means for executing the image automatic selection method.

Images