JP6245968B2 - Image processing apparatus, imaging apparatus, and control method - Google Patents

Description

  The present invention relates to an image processing device, an imaging device, and a control method.

  When shooting a subject with an image pickup device, in order to correct a shake of a picked-up image caused by a camera shake of a photographer, a camera shake correction that detects an acceleration sensor (gyro) or a motion vector of the image and corrects a shake of the picked-up image is performed. Proposed. This camera shake correction process is generally called an image stabilization process. As an example of the image stabilization process, Patent Document 1 discloses a shake correction apparatus that implements an image stabilization process using an amount of angular blur, a motion vector, and its reliability that are generated in an imaging device calculated from a gyroscope.

JP 2002-16836 A

  In order to achieve image stabilization processing within the imaging apparatus, the imaging apparatus preferably includes a motion vector detection circuit that detects a motion vector with high quality while satisfying requirements such as a low circuit scale, low cost, and real-time computation. . However, it is difficult to mount a motion vector detection circuit that satisfies all the requirements in an imaging apparatus. Therefore, depending on the characteristics of the subject to be shot, shooting conditions, shooting posture, etc., it is not always possible to calculate a high-quality motion vector for all frames to be shot. An accurate anti-vibration result could not be obtained.

  An object of the present invention is to provide an image processing apparatus that realizes a highly accurate image stabilization process even when a low-quality motion vector is detected during imaging.

  An image processing apparatus according to an embodiment of the present invention includes: a first detection unit that detects a first motion vector of an image included in a captured moving image when the moving image is captured; and the first detection unit. In order to reduce the shake in the moving image based on the reliability calculation means for calculating and storing the reliability of the detected first motion vector and the motion vector detected by the first detection means. Image stabilization parameter calculating means for calculating the image stabilization parameter, image stabilization processing means for performing image stabilization processing on the captured moving image based on the image stabilization parameter, and the moving image before the image stabilization processing Storage means for storing an image included in the image as a pre-vibration processing image, and the first motion based on the pre-vibration processing image stored by the storage means during a predetermined operation other than when capturing a moving image. Than vector And a second detecting means Yoriyukido detects the high second motion vector. For an image in which the reliability of the first motion vector is lower than a specified value, the image stabilization parameter calculation unit calculates the image stabilization parameter based on the second motion vector, and the image stabilization processing unit includes: Based on the image stabilization parameter, the image before image stabilization processing is subjected to image stabilization processing to generate an image after image stabilization processing.

  According to the image processing apparatus of the present invention, even when a low-quality motion vector is detected during imaging, high-accuracy image stabilization processing can be realized.

1 is a diagram illustrating a configuration example of an image processing apparatus according to a first embodiment. 1 is a diagram illustrating a configuration example of an image processing apparatus according to a first embodiment. It is a figure explaining the example of an operation process of the imaging device at the time of recording of an image. FIG. 10 is a diagram illustrating an example of operation processing performed by the imaging apparatus during image reproduction. It is a figure explaining a block matching method. It is a figure explaining the reliability of a motion vector. It is a conceptual diagram of the method of hierarchical motion vector detection. 6 is a diagram illustrating a configuration example of an imaging apparatus according to Embodiment 2. FIG. 6 is a diagram illustrating a configuration example of an imaging apparatus according to Embodiment 2. FIG. FIG. 9 is a diagram illustrating a configuration example of an image pickup apparatus according to a third embodiment when recording an image.

Example 1
1 and 2 are diagrams illustrating a configuration example of the image processing apparatus according to the first embodiment. The image processing apparatus illustrated in FIGS. 1 and 2 is an imaging apparatus.

FIG. 1 shows a configuration related to processing at the time of imaging among the configurations of the image processing apparatus. Specifically, FIG. 1 shows a configuration relating to storage processing of an image included in a captured moving image.
In FIG. 1A, 101 is a lens, 102 is an analog electrical signal acquired from a sensor that converts an optical signal input through the lens into an analog electrical signal, and sensor correction, noise removal, development processing, etc. A camera signal processing unit for

  Reference numeral 103 denotes a motion vector detection unit. The motion vector detection unit 103 functions as a first detection unit that detects a motion vector (first motion vector) by performing a correlation operation between the current frame and the past frame on the output of the camera signal processing unit. There is a block matching method as a representative method of motion vector detection.

  FIG. 5 is a diagram for explaining the block matching method. First, a plurality of N × N block reference blocks 302 are arranged for the current frame 301. Next, with respect to the arbitrary reference block 313 of the current frame 312, correlation calculation is performed on pixels in the N × N block existing in the search range (N + n) × (N + n) in the past frame 311, and the position with the highest degree of correlation is obtained. Is a motion vector. In the example shown in FIG. 5, since the N × N block at the position 315 has the highest correlation, 314 is detected as a motion vector. Note that means for calculating an evaluation value that is a correlation value include a sum of absolute differences, a sum of squared differences, a normal cross-correlation value, and the like for pixels in an N × N block. Here, when the difference absolute value sum is an evaluation value, the correlation is highest when the difference absolute value sum is the lowest.

FIG. 6 is a diagram for explaining the reliability of the motion vector.
The motion vector detection unit 103 performs a motion vector detection process and also functions as a reliability calculation unit that calculates the reliability of the motion vector. In each graph 401 to 404 in FIG. 6, the horizontal axis represents the coordinate position of the search range in the horizontal direction with respect to the reference block, and the vertical axis represents the sum of absolute differences at each coordinate position. Although FIG. 6 shows a two-dimensional display in the horizontal direction, it is actually expressed as a three-dimensional distribution in consideration of the vertical direction.

  In the graph 401, it can be said that the reliability of the motion vector is high because the difference absolute value sum has one minimum value and is sufficiently smaller than the other difference absolute value sums. On the other hand, in the graph 402, there is a minimum value of the sum of absolute differences, but since all the sums of absolute differences are high, a highly correlated block could not be detected, so it can be said that the reliability of the motion vector is low. In addition, in the graph 403, there is a minimum value of the sum of absolute differences, but since the values of all the sums of absolute differences are low, it can be estimated that the motion vector is a flat part and a low contrast part that are difficult to detect. Furthermore, in the graph 404, since there are a plurality of minimum values of the sum of absolute differences, it can be estimated that the motion vector is a striped pattern, lattice pattern, or line that is difficult to detect. In this way, the reliability of the detected motion vector is calculated using the sum of absolute differences calculated at the time of detecting the motion vector.

  The motion vector reliability information calculation is not limited to the method described above. For example, when the imaging device is equipped with a gyro sensor, the degree of divergence between the angle blur amount and the motion vector obtained from the gyro sensor output is used as the reliability, and the method is used in which the reliability is low when the deviation is large. Also good. It is assumed that information such as a frame number is given to the reliability information.

  The image stabilization parameter calculation unit 104 calculates the angle shake amount of the image pickup device from the integrated value of the output of the gyro sensor provided in the image pickup device, and reduces the shaking in the moving image based on the angle shake amount and the motion vector. A vibration isolation parameter is calculated. The anti-vibration parameter here is specifically a projective transformation matrix or an affine transformation matrix for the geometric deformation unit 105. The geometric deformation unit 105 performs image stabilization processing on the image output from the camera signal processing unit 102 using the image stabilization parameter calculated by the image stabilization parameter calculation unit 104.

  The reduction unit 106 reduces the image subjected to image stabilization processing by the geometric deformation unit 105 so as to match the number of pixels of the display unit. The display unit 107 displays the image reduced by the reduction unit 106. The display unit is, for example, EVF. The image after the image stabilization process displayed on the display unit 107 (hereinafter referred to as “image after image stabilization process”) is subjected to a high-accuracy image stabilization process when the reliability of the motion vector is low. It is not displayed, and an image with a shake remaining or an angle of view is displayed.

  Reference numeral 108 denotes an image storage unit before image stabilization processing that stores an image before image stabilization processing (hereinafter referred to as “image before image stabilization processing”). Reference numeral 109 denotes a motion vector storage unit that stores the motion vector detected by the motion vector detection unit 103. Reference numeral 110 denotes a reliability information storage unit that stores motion vector reliability information calculated when the motion vector detection unit 103 detects a motion vector.

FIG. 3 is a flowchart for explaining an operation processing example of the imaging apparatus at the time of image recording (storage) in the first embodiment.
In S201, the recording operation of the imaging apparatus is started, and the process proceeds to S202. In S202, an image is input to the imaging apparatus through the lens 101, and the process proceeds to S203. In S203, the camera signal processing unit 102 performs image development processing, and the process proceeds to S204. In step S204, the motion vector detection unit 103 calculates the motion vector and the sum of absolute differences used when detecting the motion vector by using the current frame image and the past frame image by the method described with reference to FIG. Migrate to In S205, the motion vector detection unit 103 calculates the reliability information of the motion vector using the absolute difference sum, and proceeds to S206 and S208.

  In S206, the motion vector detection unit 103 stores the reliability information of the calculated motion vector in the reliability information storage unit 110. In S207, the imaging apparatus records the image before image stabilization processing in the image storage unit 108 before image stabilization processing, and proceeds to S211.

  On the other hand, in S208, the image stabilization parameter calculation unit 104 calculates the image stabilization parameter using the motion vector and the gyro sensor output, and the process proceeds to S209. Subsequently, in S209, the geometric deformation unit 105 performs image stabilization processing on the image using the image stabilization parameter, and the process proceeds to S210.

  After the reduction unit 106 performs the reduction process on the image after the image stabilization process, in step S210, the display unit 107 displays the image after the reduction process. As a result, the image after the image stabilization process becomes a display target, and a highly accurate image stabilization process is realized. In S211, if there is a recording end instruction for the imaging apparatus, the recording is ended. If there is no instruction, the process proceeds to S202, and the above-described processing is repeated until the recording ends.

  FIG. 2 is a diagram illustrating a configuration example related to image display processing in the configuration of the image processing apparatus. Note that the same processing units as those shown in FIG. The high-quality motion vector detection unit 120 functions as a second detection unit that detects a high-quality motion vector (second motion vector) by executing a hierarchical motion vector detection process. In this example, the high-definition motion vector detection unit 120 detects a high-quality motion vector when an image is displayed. However, the high-definition motion vector detection unit 120 does not capture an image, that is, at the time of camera recording. A high-quality motion vector may be detected during a predetermined operation other than the above. For example, the high-quality motion vector detection unit 120 may detect a high-quality motion vector during standby, charging, or the like.

  FIG. 7 is a conceptual diagram of a hierarchical motion vector detection method. Reference numeral 501 denotes a recorded image, and the image size becomes smaller as the images 502 and 503 are obtained. Although the motion vector accuracy is coarse in the lowest layer image 503, a motion vector can be detected from a wide search range, and an approximate motion can be obtained. Next, if the search is performed centering on the motion vector position calculated in the lower hierarchy in the intermediate hierarchy and the highest hierarchy, the motion vector accuracy can be improved while narrowing down the search range and narrowing down the candidates. That is, the high-quality motion vector detection unit 120 detects the second motion vector based on images of a plurality of hierarchies having different image sizes and motion vector search ranges corresponding to the pre-vibration processing image.

  The means for calculating a high-quality motion vector is not limited to the method described with reference to FIG. For example, a motion vector may be calculated from the correspondence between feature points such as a corner of a subject using a CPU or dedicated hardware that exists outside an imaging apparatus such as a server or a smartphone.

  Returning to the description of FIG. The switch 121 uses the motion vector reliability information to select which of the motion vector stored in the motion vector storage unit 109 and the high quality motion vector detected by the high quality motion vector detection unit 120 is to be selected. Switch based on. When the reliability of the motion vector stored in the motion vector storage unit 109 is equal to or higher than the specified value, the switch 121 selects the motion vector (the motion vector recorded at the time of image recording) and selects the image stabilization parameter calculation unit 104. Output to. When the reliability of the motion vector stored in the motion vector storage unit 109 is lower than the specified value, the switch 121 selects the high-quality motion vector detected by the high-quality motion vector detection unit 120 and calculates the image stabilization parameter. Output to the unit 104.

  The geometric deformation unit 105 performs image stabilization processing using the new image stabilization parameter calculated by the image stabilization parameter calculation unit 104, and records (stores) the image after image stabilization processing in the image storage unit 122 after image stabilization processing. .

FIG. 4 is a flowchart illustrating an example of operation processing performed by the imaging apparatus during image reproduction according to the first embodiment. This flowchart is executed in accordance with control of a CPU (not shown) included in the imaging apparatus.
In S221, the reproduction operation of the imaging apparatus is started, and the process proceeds to S222. The imaging apparatus inputs the pre-vibration image recorded in the pre-vibration image storage unit to the high-quality motion vector detection unit 120, and the process proceeds to S223. In S223, the imaging apparatus inputs motion vector reliability information from the reliability information storage unit 110 to the switch 121, and the process proceeds to S224. The input motion vector reliability information is the reliability of the motion vector of the current frame calculated at the time of image recording.

  In S224, the switch 121 determines whether the reliability of the motion vector of the current frame calculated at the time of recording is low. If the reliability of the motion vector of the current frame calculated at the time of recording is low, the process proceeds to S225. When the reliability of the motion vector of the current frame calculated at the time of image recording is high (greater than a specified value), the process proceeds to S226.

  In S225, the high quality motion vector detection unit 120 calculates a high quality motion vector. Then, the switch 121 inputs a high-quality motion vector to the image stabilization parameter calculation unit 104, and the process proceeds to S227. In S226, the switch 121 inputs the recorded motion vector of the current frame from the motion vector storage unit 109 to the image stabilization parameter calculation unit 104, and proceeds to S227.

  In S227, the image stabilization parameter calculation unit 104 calculates the image stabilization parameter based on the motion vector input from the switch 121. When the output information of the gyro sensor is recorded at the time of recording, the image stabilization parameter calculation unit 104 may calculate the image stabilization parameter using the output information.

  In S228, the geometric deformation unit 105 performs a geometric deformation process (anti-vibration process) using the anti-vibration parameter calculated by the anti-vibration parameter calculation unit 104, and generates an image after anti-vibration process. Then, the process proceeds to S229. In S229, the geometric deformation unit 105 records the image after the image stabilization process in the image storage unit 122 after the image stabilization process. In S230, the display unit 107 displays the image after the image stabilization process.

  In S231, when there is an instruction to end the reproduction to the imaging apparatus, or when reproduction is performed up to the last frame of the data to be reproduced, the reproduction is terminated, and when there is no instruction and reproduction is not performed up to the final frame, the process proceeds to S222. Then, the above-described processing is repeated until the end of reproduction.

  The operation at the time of reproduction may be performed as background processing not only when a reproduction instruction is given to the image pickup apparatus but also when the image pickup apparatus is not used at the time of charging or the like. At this time, since it is not necessary to display a vibration-proof image on the display unit, output to the display unit is not performed. Furthermore, this operation can be performed a plurality of times.

(Example 2)
The imaging apparatus according to the second embodiment records a pre-vibration processing image and a post-vibration processing image for a frame with low motion vector reliability, calculates a motion vector for each image, and selects a higher quality image.

  8 and 9 show a configuration example of the imaging apparatus according to the second embodiment. FIG. 8 shows a configuration during recording, and FIG. 9 shows a configuration during reproduction.

  In FIG. 8, reference numeral 601 denotes a lens, and reference numeral 602 denotes a camera that digitizes an analog electrical signal acquired from a sensor that converts an optical signal input through the lens into an analog electrical signal to perform sensor correction, noise removal, development processing, and the like. It is a signal processing unit.

  As in the first embodiment, reference numeral 603 denotes a motion vector detection unit that performs a correlation operation between the current frame and the past frame on the output of the camera signal processing unit, and calculates the reliability of the motion vector and the motion vector.

  Reference numeral 604 denotes an image stabilization parameter calculation unit. When the image pickup apparatus includes a gyro sensor, the image stabilization parameter calculation unit 604 calculates the angle shake amount of the image pickup apparatus from the integral value of the gyro sensor output, and suppresses the shaking together with the motion vector. Calculate anti-vibration parameters. Reference numeral 605 denotes a geometric deformation unit. The geometric deformation unit 605 executes a geometric deformation process (anti-vibration process) on the image from the camera signal processing unit 602 based on the image stabilization parameter calculated by the image stabilization parameter calculation unit 604. Reference numeral 606 denotes a reduction unit that reduces the image after image stabilization processed by the geometric deformation unit 605 so as to match the number of pixels of the display unit. The display unit 607 displays the reduced image.

  Reference numeral 608 denotes a pre-vibration processing image storage unit that stores a pre-vibration processing image. Reference numeral 609 denotes a motion vector storage unit that stores the motion vector calculated by the motion vector detection unit 603. A reliability information storage unit 610 stores motion vector reliability information calculated when the motion vector detection unit 603 detects a motion vector.

  Reference numeral 611 denotes a switch. Based on the reliability of the motion vector stored in the reliability information storage unit 610, the switch 611 determines whether or not to record the image after image stabilization processed by the image pickup apparatus in the image storage unit 612 after image stabilization processing. decide. When the reliability of the motion vector is low, the image after the image stabilization process is recorded. When the reliability of the motion vector is high, the image after the image stabilization process is not recorded.

  FIG. 9 will be described. Reference numerals 620 and 621 denote high-quality motion vector detection units. The high-quality motion vector detection unit 620 detects a motion vector using the pre-vibration processing image stored in the pre-vibration processing image storage unit 608. The high-quality motion vector detection unit 621 acquires an image after image stabilization processing during image recording from the image storage unit 612 after image stabilization processing, and detects a motion vector using the acquired image after image stabilization processing. The specific detection method is hierarchical motion vector detection or motion vector detection using an external CPU or external dedicated hardware, as in the first embodiment.

  Reference numeral 622 denotes a switch, and 623 denotes a reliability determination unit. The reliability determination unit 623 determines which of the motion vectors output from each of the high-quality motion vector detection units 620 and 621 is high in reliability, and selects a motion vector with high reliability. The switch 622 is instructed to output.

  The switch 622 acquires the reliability of the recorded motion vector from the reliability information storage unit 610. When the reliability of the recorded motion vector is low, the high-quality motion vector selected by the switch 622 is selected and output. When the reliability of the recorded motion vector is high, the recorded motion vector is output.

  The image stabilization parameter calculation unit 604 calculates the image stabilization parameter using the motion vector output from the switch 624, and the geometric deformation unit 605 executes the image stabilization process. The geometric deformation unit 605 stores the image after the image stabilization process in the image storage unit 625 after the image stabilization process. Further, the display unit 607 displays the image after the image stabilization process.

(Example 3)
The imaging apparatus according to the third exemplary embodiment predicts whether or not a frame with low motion vector reliability occurs (motion vector error occurs). When predicting that a motion vector error occurs, emphasize the feature of the image, Images that are easy to detect high-quality motion vectors are also recorded.

  FIG. 10 is a diagram illustrating a configuration example of the imaging apparatus according to the third embodiment at the time of image recording. The configuration at the time of image reproduction is the same as the configuration at the time of reproduction in the first embodiment or the second embodiment, and thus description thereof is omitted.

  In FIG. 10, the lenses 701 to 706 are the same as the lenses 601 to 607 in FIG. Further, the pre-vibration processing image storage unit 708 to the motion vector storage unit 706 are the same as the pre-vibration processing image storage unit 608 to the motion vector storage unit 709 of FIG.

  Reference numeral 711 denotes a motion vector error occurrence prediction unit. The motion vector error occurrence prediction unit 711 predicts whether or not a motion vector error occurs. The motion vector error occurrence prediction unit 711 is at least one of a case where the photographing posture is a walk on the tele side, a case where the camera work is large, and a case where the amount of angular blur calculated by the gyro is equal to or larger than a predetermined amount. In addition, a motion vector error is predicted to occur. The motion vector error occurrence prediction unit 711 outputs a prediction result (prediction information) of occurrence of a motion vector error. This prediction information is input to the switch 713.

  Reference numeral 712 denotes a filter processing unit. The filter processing unit 712 executes enhancement processing (filter processing) that enhances the image characteristics with a high-pass filter that enhances high-frequency components such as edges so that a motion vector can be easily detected. Output.

  The switch 713 stores the filtered image in the pre-vibration processing image storage unit 714 for a frame in which a motion vector error is predicted to occur based on the input prediction information.

  At the time of image reproduction, the pre-vibration image storage units 714 and 708 function as, for example, the pre-vibration image storage unit 608 in FIG. Therefore, the high-quality motion vector detection unit 620 detects a motion vector based on the image acquired from the pre-vibration processing image storage unit 714 for a frame in which a motion vector error is predicted to occur.

(Other examples)
The present invention can also be realized by executing the following processing. That is, software (program) that realizes the functions of the above-described embodiments is supplied to a system or apparatus via a network or various storage media, and a computer (or CPU, MPU, or the like) of the system or apparatus reads the program. It is a process to be executed. In this case, the program and the storage medium storing the program constitute the present invention.

DESCRIPTION OF SYMBOLS 101 Lens 102 Camera signal processing part 103 Motion vector detection part 104 Anti-vibration parameter calculation part 105 Geometric deformation part 106 Reduction part 107 Display part

Claims (9)

First detection means for detecting a first motion vector of an image included in the captured moving image when capturing a moving image;
Reliability calculation means for calculating and storing the reliability of the first motion vector detected by the first detection means;
Anti-vibration parameter calculating means for calculating an anti-shake parameter for reducing shake in the moving image based on the motion vector detected by the first detecting means;
Anti-vibration processing means for performing anti-vibration processing on the captured moving image based on the anti-vibration parameters;
Storage means for storing an image included in the moving image before the image stabilization process as an image before the image stabilization process;
A second motion vector for detecting a second motion vector having a higher reliability than the first motion vector based on the pre-vibration processing image stored by the storage means during a predetermined operation other than when capturing a moving image; Detecting means,
For an image in which the reliability of the first motion vector is lower than a specified value, the image stabilization parameter calculation unit calculates the image stabilization parameter based on the second motion vector, and the image stabilization processing unit includes: An image processing apparatus, comprising: generating an image after image stabilization processing by performing image stabilization processing on the image before image stabilization processing based on the image stabilization parameter. For an image in which the reliability of the first motion vector is higher than a specified value, the image stabilization parameter calculation unit calculates the image stabilization parameter based on the first motion vector, and the image stabilization processing unit includes: The image processing apparatus according to claim 1, wherein the image before the image stabilization process is subjected to an image stabilization process based on the image stabilization parameter. At the time of capturing the moving image, for an image in which the reliability of the first motion vector is lower than a specified value, the image stabilization processing unit performs image stabilization processing and stores the image as an image after image stabilization processing,
During a predetermined operation other than the time of capturing a moving image, the second detection unit, for an image whose reliability of the first motion vector is lower than a specified value, the image before image stabilization processing and the image after image stabilization processing The image processing apparatus according to claim 1, wherein a motion vector having a higher reliability among motion vectors calculated based on each of the motion vectors is detected as the second motion vector. For an image in which the occurrence of an error in which the reliability of the first motion vector is lower than a specified value is predicted when the moving image is captured, the storage unit prevents the image in which the high-frequency component of the image is emphasized. The image processing apparatus according to claim 1, wherein the image processing apparatus stores the pre-vibration image. Predicting means for predicting the occurrence of the error and outputting prediction information of the occurrence of the error;
An enhancement processing means for performing processing for enhancing a high-frequency component of an image included in the captured moving image;
The storage means determines whether the error occurs based on the prediction information of the occurrence of the error, and for the image determined to generate the error, the image obtained by the enhancement process is used as the image stabilization process. The image processing apparatus according to claim 4, wherein the image processing apparatus is stored as a previous image. The prediction means is at least when the imaging posture of the imaging device that captures the moving image is a walk on the tele side, when the camera work is large, or when the amount of angular blur of the imaging device is greater than or equal to a specified value The image processing apparatus according to claim 5, wherein the occurrence of the error is predicted in any of the cases. The second detection means detects the second motion vector based on images of a plurality of layers having different image sizes and motion vector search ranges corresponding to the image before image stabilization processing. The image processing apparatus according to any one of claims 1 to 6.   An imaging apparatus that functions as the image processing apparatus according to claim 1. A first detection step of detecting a first motion vector of an image included in the captured moving image when capturing a moving image;
A reliability calculation step of calculating and storing the reliability of the first motion vector detected by the first detection step;
An image stabilization parameter calculating step for calculating an image stabilization parameter for reducing shake in the moving image based on the motion vector detected by the first detection step;
Based on the anti-vibration parameters, anti-vibration processing step of performing anti-vibration processing on the captured moving image,
A storage step of storing an image included in the moving image before the image stabilization process as an image before the image stabilization process;
A second detection step of detecting a second motion vector having higher reliability than the first motion vector based on the stored pre-vibration processing image during a predetermined operation other than when capturing a moving image; And
For an image in which the reliability of the first motion vector is lower than a specified value, the image stabilization parameter calculation step calculates the image stabilization parameter based on the second motion vector, and the image stabilization processing step A control method for an image processing apparatus, comprising: generating an image after image stabilization processing by performing image stabilization processing on the image before image stabilization processing based on the image stabilization parameter.

Images