JP5598569B2 - Panorama photographing apparatus and method, camera unit equipped with panoramic photographing apparatus - Google Patents

Description

  The present invention can shoot a panoramic image with a large angle of view even in an electronic device including a camera module composed of a lens module with a small angle of view by capturing the image in a plurality of times and combining the captured images. It relates to technology.

  A camera module for a small electronic device with a digital camera generally uses a small lens module in order to prioritize miniaturization, and can only capture images with a small angle of view. When using a digital camera equipped with such a small lens module to capture a wide-angle subject, such as a landscape photograph, multiple partially overlapping images taken in multiple shots This is realized by combining the images.

  Since it is necessary to specify the position to be combined in the partially overlapping images, the photographer must follow the guide displayed on the viewfinder screen and align the device in the correct direction when shooting multiple images. By repeating the photographing operation, a plurality of images to be combined are obtained.

  Japanese Patent Application Laid-Open No. 2004-228561 discloses a technique for combining a plurality of images based on a set alignment amount and a joint portion between a plurality of partially overlapping images.

JP 2004-236910 A

  With conventional technology, images can be taken multiple times, and the captured images can be combined to capture images with a large angle of view even in electronic devices with a camera module consisting of a lens module with a small angle of view. The photographer needs to repeat the shooting operation after aligning the device in the correct direction according to the guide displayed on the viewfinder screen, and the operation is complicated. In addition, when a photographer who is unfamiliar with the operation fails to align the apparatus in the correct direction, a combined image in which the image of the joint portion is shifted may be obtained.

  Also, it is conceivable to synthesize panoramic images from continuous images taken as moving images, but since various moving people or objects enter the continuous images of moving images, simply combining adjacent images There is a high possibility that the panorama image cannot be synthesized well.

  It is an object of the present invention to be able to shoot with a simplified procedure when shooting an image to be combined, to reduce the burden on the photographer, and to become a combined image shifted by a photographer who is unfamiliar with the operation. It is to reduce the failure of the shooting procedure.

The present invention is premised on an apparatus or method for synthesizing panoramic images from continuous images captured by a camera module, and a camera unit equipped with the panoramic imaging apparatus.
The following will describe aspects when the present invention is realized as a panoramic photographing apparatus.

The motion vector calculation means (201) acquires position difference information of the continuous images being shot.
The motion vector quality determination means (202) determines the usefulness of the position difference information acquired by the motion vector calculation means.

  The motion amount calculation means (203) calculates the motion amount of the image when the motion vector quality determination means determines that the motion amount of the image can be calculated. For example, the motion amount calculation means calculates a motion amount between images based on a motion vector for each macroblock of a continuous image input from a camera module. In addition, the motion amount calculation unit obtains an average value of motion vectors for each macroblock, for example, when calculating the motion amount, and a motion vector having a certain error or more with respect to the average value is a target of motion amount calculation. Exclude more. Further, the motion amount calculation means calculates the motion amount by, for example, reusing the motion amount calculated in the previous image when there are few effective motion vectors and the motion amount cannot be calculated. Compensate for the amount of movement that was impossible.

  The image selection means (205) selects an image to be processed based on the amount of movement. For example, the image selection unit selects an image to be combined when a certain amount of motion calculated by the motion amount calculation unit is accumulated.

  The image combining means (206) combines the overlapping portions using a plurality of partially overlapping images selected by the image selecting means and combines them into one image.

  According to the present invention, when taking an image to be combined, it is possible to shoot with a more simplified procedure, reducing the burden on the photographer, and a combined image shifted by a photographer unfamiliar with the operation It is possible to reduce the failure of the shooting procedure.

  In particular, it is possible to synthesize an optimal panoramic image from a continuous image captured as a moving image while appropriately processing people / objects that move in various ways through motion vectors.

It is a figure which shows the structure of the electronic device provided with the digital camera module and image processor by embodiment of this invention. It is a figure which shows imaging | photography operation of an electronic device provided with the digital camera module and image processing processor by embodiment of this invention. It is a figure shown about the input image and output result of a motion vector calculation part. It is an operation | movement flowchart for determining whether the motion amount of an image is computable performed by the motion vector quality determination part. It is an operation | movement flowchart for calculating the motion amount of an image performed by the motion amount calculation part. It is an operation | movement flowchart for calculating the motion amount of an image performed by the motion amount calculation complement part. It is an operation | movement flowchart for selecting the image used as a joint target from the continuous image, and calculating the positional relationship between a joint target image performed by the image selection part. Motion amount calculation unit, motion amount calculation complementing unit, image selection unit, and image combination unit are executed to calculate the amount of motion of continuous images to be executed, to select images to be combined, and to accumulate motion amount It is operation | movement explanatory drawing which shows calculation of this, and the coupling | bonding of an image.

The best mode for carrying out the present invention will be described below in detail with reference to the drawings.
In the drawings, similar components are given the same reference numerals.
FIG. 1 shows a configuration of an electronic apparatus 1 including a camera module 10 according to the present embodiment. The electronic device 1 further includes an image processing processor (CPU) 20, a recording unit 30, and a user interface (I / F) 40 coupled to an input device 401 (for example, a key) and a display 402.

  The camera module 10 includes a lens module 101, a CCD / CMOS sensor 102 for imaging, an analog-digital converter (ADC) 103, and a digital signal processor (DSP) 104.

  The image processor 20 calculates a motion amount by selecting a motion vector calculation unit 201 for calculating a position difference with respect to the previous image for each small rectangular area called a macroblock in the image, and selecting a motion vector to be calculated for image selection. Motion vector quality determination unit 202 that determines whether or not it is possible, motion amount calculation unit 203 that calculates the amount of motion of an image from motion vector information, and past calculation history when information necessary for motion amount calculation is not obtained A motion amount calculation complementing unit 204 that estimates a motion amount, an image selection unit 205 that determines a motion amount accumulated value and a combination target image from the calculated motion amount, and an overlapping portion of the combination target images are joined to form one image The image combining unit 206 is combined with the image combining unit 206. For example, at least a part of the image processor 20 may be implemented in the form of hardware as an integrated circuit, or at least a part of the image processor 20 may be implemented in the form of software as a program operating on the processor.

The recording unit 30 includes a continuous image area 301, a selected image and accumulated motion amount value area 302, and a combined image area 303. The continuous image area 301 stores a plurality of images acquired by the camera module 10 at regular time intervals. The selected image and accumulated motion amount value area 302 stores a plurality of images to be combined and a accumulated motion amount value indicating a relative position between the images. The combined image area 303 stores the output image that has undergone image processing.

  The user interface 40 is coupled to an input device 401 including a key and a display 402, supplies a photographer's key input to the camera module 10, and displays a captured image and a processed image on the display 402.

  In FIG. 1, the camera module 10 starts continuous image capturing in response to the photographer pressing the image capturing start button, and stores the image data in a continuous image area 301. A photographer inputs an image of a wide-angle photographing object such as a landscape photograph to the camera module 10 by translating the electronic apparatus 1 during photographing of continuous images. The motion vector calculation unit 201 calculates position difference information from the previous image called a motion vector for each 16 × 16 pixel section called a macroblock in the image for the input continuous image. The motion vector quality determination unit 202 selects only the motion vectors necessary for processing from the motion vectors of all macroblocks, and determines whether the motion amount can be calculated. The motion amount calculation unit 203 calculates the motion amount of the image based on the motion vector selected by the motion vector quality determination unit 202. When the number of effective motion vectors is small as a result of the selection, the motion amount calculation complementing unit 204 performs processing using the motion amount calculated in the past. The image selection unit 205 accumulates the motion amount, and selects an image at the time when a certain amount of motion is accumulated as a combination target. The image combining unit 206 combines the selected images to be combined, and stores a wide-angle image as an output result in the combined image area 303.

  FIG. 2 illustrates a photographing operation of the electronic apparatus 1 according to the present embodiment. In response to the photographer pressing the shooting start button, the electronic device 1 starts shooting a plurality of continuous images, and stores a plurality of image data captured in the continuous image region 301. The photographer inputs an image of a wide-angle shooting target such as a landscape photograph to the camera module 10 by translating the electronic device 1 during continuous image shooting. In the figure, (a) to (c) respectively indicate the position of the electronic device 1 at the start of shooting, the position of the electronic device 1 during shooting, and the position of the electronic device 1 at the end of shooting.

  FIG. 3 illustrates the input image and the output result of the motion vector calculation unit 201. The motion vector calculation unit is based on the input of the continuous image from the continuous image region 301, and the previous image (d) and the current image (e) having a two-dimensional value called a motion vector for each 16 × 16 pixel section called a macroblock. And position difference information is calculated. In some macroblocks, a motion vector cannot be calculated, and a macroblock incapable of calculating a motion vector is called an intra macroblock. (f) shows an image partitioned into macroblocks and a motion vector for each macroblock.

  FIG. 4 is a flowchart for determining whether or not the motion amount of an image can be calculated, which is executed by the motion vector quality determination unit 202 of the image processor 20 according to the present embodiment.

  In steps S501 to S502, the motion vector quality determination unit 202 reads a motion vector for each macroblock in the image obtained as a result of the processing of the motion vector calculation unit 201. Some macroblocks may be intra macroblocks that do not have motion vectors, and moving subjects that are moving during shooting may frame in, so the macroblocks and intra macroblocks at the top, bottom, left, and right edges of the image are It is not used for calculating the amount of motion of an image. In step S503, it is determined whether there are a predetermined number or more of macroblocks necessary for calculating the motion amount. If it is determined in step S503 that there are no more than a certain number of macroblocks necessary for calculating the motion amount, it is determined in step S508 that the motion amount of the target image cannot be calculated.

  If it is determined in step S503 that there is a certain number or more of macroblocks necessary for calculating the amount of motion, it is determined that the amount of motion of the target image can be calculated, and in steps S504 to S505, the top, bottom, left, and right edges of the image. The average value μ and the variance value ρ of the motion vectors of the macroblocks other than the non-intra macroblock are calculated. For macroblocks with motion vectors that exceed the average value μ to ± ρ, it is determined that this is due to noise in the video or a moving subject that is moving during shooting. Not used for calculation. In step S506, it is determined whether or not there are a certain number or more of macroblocks whose motion vectors are in the range of μ ± ρ, which are targets of motion amount calculation. If it is determined in step S506 that there are no more than a certain number of macroblocks necessary for calculating the motion amount, it is determined in step S508 that the motion amount of the target image cannot be calculated. If it is determined in step S506 that there is a certain number or more of macroblocks necessary for calculating the motion amount, it is determined in step S507 that the motion amount of the target image can be calculated.

FIG. 5 shows a flowchart for calculating the motion amount of the image, which is executed by the motion amount calculation unit 203 of the image processor 20 according to the present embodiment.
When the motion vector quality determination unit 202 determines that the motion amount of the image can be calculated, the motion amount detection unit 203 calculates the motion amount of the image. In step S511, the motion amount detection unit 203 reads a motion vector for each macroblock in the image. In step S511, depending on the macro block, there may be an intra macro block having no motion vector, and a moving shooting target may be framed in during shooting. Intra macroblocks are not used to calculate the amount of motion of an image. In step S512, an average value μ and a variance value ρ of macroblocks other than the top, bottom, left, and right edges in the image and not an intra macroblock are calculated. For macroblocks with motion vectors that exceed the average value μ to ± ρ, it is determined that this is due to noise in the video or a moving subject that is moving during shooting. Not used for calculation. In steps S513 to S514, the average value μ ′ of the motion vectors of the macroblocks whose motion vectors are in the range of μ ± ρ is calculated, and μ ′ is set as the motion amount of the target image.

FIG. 6 shows a flowchart for calculating the motion amount of the image, which is executed by the motion amount calculation complementing unit 204 of the image processor 20 according to this embodiment.
When the motion vector quality determination unit 202 determines that the image motion amount cannot be calculated, the motion amount calculation complementing unit 204 calculates the image motion amount. Even when shooting a subject whose motion amount cannot be calculated at the time of shooting, the photographer assumes that the electronic device 1 is translated at a constant speed and uses the motion amount calculated in the previous image. To continue processing. In steps S520 to S521, the motion amount calculated by the motion amount detection unit 203 or the motion amount calculation supplement unit 204 in the previous image is read and applied as the motion amount of the current image.

FIG. 7 shows a flowchart for selecting the images to be combined from the continuous images and calculating the positional relationship between the images to be combined, executed by the image selection unit 205 of the image processor 20 according to the present embodiment. ing.

  The image selection unit 205 selects the images to be combined from the continuous images, and the positional relationship between the images to be combined based on the motion amounts calculated by the motion amount calculation unit 203 and the motion amount calculation complement unit 204. Is calculated. In step S530, the head image of the continuous image is acquired. In step S531, the accumulated motion amount is initialized. The first image is selected as a combination target image in step S537. For the subsequent images, the motion amount cumulative value is initialized in step S532. In steps S533 to S535, the next image and the amount of motion of the image are acquired, and the amount of motion of the image is added to the accumulated amount of motion. In steps S536 to S537, an image at the time when the accumulated motion amount becomes equal to or greater than the predetermined value Vs is selected as a combination target. Vs corresponds to the difference between the size of the images to be combined and the size of the partially overlapped area between the images necessary for combining the images. If the accumulated motion amount is less than Vs, the process returns to step S533 again to acquire the next image. If it is determined in step S538 that more than the expected number of images have been selected, the process is completed.

  FIG. 8 shows the calculation of the motion amount of the continuous image executed by the motion amount calculation unit 203 and the motion amount calculation complement unit 204 of the image processor 20 and the combination executed by the image selection unit 205 according to this embodiment. FIG. 10 illustrates the selection of an image to be processed, the calculation of a cumulative amount of motion, and the image combination performed by the image combining unit 206.

  In the figure, (g) illustrates the motion amount calculation result of the continuous image executed by the continuous image and motion amount calculation unit 203 and the motion amount calculation complement unit 204. In the figure, (h) to (j) illustrate the selection results of the images to be combined, which are executed by the image selection unit 205. In the drawing, (k) illustrates the calculation result of the accumulated motion amount executed by the image selection unit 205. In the drawing, (l) illustrates a combined image that is a result of combining images executed by the image combining unit 206.

  The embodiments described above are merely given as typical examples, and it is obvious to those skilled in the art to combine the components of each embodiment, and variations and variations thereof will be apparent to those skilled in the art. Obviously, various modifications may be made to the above-described embodiments without departing from the scope of the invention as set forth in the scope.

DESCRIPTION OF SYMBOLS 1 Electronic device 20 Image processor 30 Recording part 40 User interface 201 Motion vector calculation part 202 Motion vector quality determination part 203 Motion amount calculation part 204 Motion amount calculation complement part 205 Image selection part 206 Image combination part

Claims (4)

A device for synthesizing panoramic images from continuous images taken by a camera module,
Motion vector calculating means for acquiring position difference information for each macroblock of the continuous image;
Motion vector quality determination means for determining the usefulness of the position difference information acquired by the motion vector calculation means based on the number of macroblocks of the continuous image that can be used for calculating the amount of motion of the image ;
A motion amount calculating means for calculating the motion amount of the image when the motion vector quality determining means determines that the motion amount of the image can be calculated;
Image selection means for selecting an image to be processed based on the amount of movement;
Image combining means for combining overlapping portions using a plurality of partially overlapping images selected by the image selecting means and combining them into one image;
A panoramic photographing apparatus characterized by comprising: The motion amount calculating means calculates a motion amount between images based on a motion vector for each macroblock of the continuous image,
The image selecting means selects the image to be combined when the certain amount of motion calculated by the motion amount calculating means is accumulated;
The panorama photographing apparatus according to claim 1. When calculating the amount of motion, the motion amount calculation means obtains an average value of motion vectors for each macroblock, and excludes motion vectors having a certain error from the average value from motion amount calculation targets. ,
The panorama photographing apparatus according to claim 1. When calculating the motion amount, the motion amount calculation means cannot calculate the motion amount by reusing the motion amount calculated in the previous image when there are few effective motion vectors and the motion amount cannot be calculated. To compensate for the amount of movement
The panorama photographing apparatus according to claim 1.