Evaluation method, evaluation device, and program

Abstract

To provide an evaluation method for accurately determining shake correction performance of an imaging apparatus.SOLUTION: A method for evaluating a shake correction effect of an imaging apparatus includes: an acquisition step of acquiring a first change amount for each photographing from a predetermined photographing number of subject images photographed by a shaken imaging apparatus; a calculation step of dividing the predetermined photographing number into a plurality of sections, and executing statistical processing on the first change amount in the plurality of sections to calculate a second change amount in each section; a determination step of determining the largest second change amount of the second change amounts in the sections as a third change amount; and an evaluation step of evaluating the shake correction effect based on the third change amount.SELECTED DRAWING: Figure 6

Description

[Technical Field]

[0001] The present invention relates to a method for evaluating an imaging device, and more particularly to a method, an evaluation device, and a program for evaluating the image stabilization effect of an imaging device. [Background technology]

[0002] As a method for evaluating the image stabilization effect or image stabilization performance of an imaging device, Patent Document 1 discloses a method for calculating an evaluation value of the image stabilization effect by taking into account the influence of the blur offset amount. Note that the blur offset amount is the amount of blur in a captured image caused by factors other than camera shake, and is a value specific to the device under test that is affected by the optical performance, number of effective pixels, image processing, etc. of the imaging device. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Patent No. 5909686 Summary of the Invention [Problem to be solved by the invention]

[0004] Patent Document 1 discloses a method for evaluating image stabilization effectiveness by observing the degree of degradation of images captured while vibrating an imaging device. Generally, when evaluating image stabilization effectiveness using a vibration generator, vibration is applied using a vibration generator with a standard-defined shake waveform and predetermined performance, and the results are compared after removing the blur offset amount with the image stabilization function OFF and with the image stabilization function ON. The number of images required to evaluate image stabilization effectiveness is said to be 200 or more. Image stabilization performance evaluation is based on the average amount of blur for the number of images captured with the imaging device.

[0005] However, the amount of shake may not be constant from the start to the end of the number of shots taken, and for example, the shake may be large in the early stages of shooting, or in the latter stages of shooting, such as when the number of shots approaches 200. When the amount of shake varies over time in this way, it is not possible to appropriately evaluate the shake correction performance by simply using the average value of all shots taken.

[0006] The present invention has been made in view of the above circumstances, and has as its object to provide an evaluation method for more accurately determining the image stabilization performance of an imaging device. [Means for solving the problem]

[0007] An evaluation method according to one embodiment of the present invention is a method for evaluating the blur correction effect of an imaging device, which evaluates a predetermined number of subject images taken by an imaging device that is vibrated. Each of mosquito Photo Shadow by Shadow 1 Shake an acquisition step of acquiring the amount of the predetermined number of images; Subject image of , so that each of the multiple sections includes multiple captured images in the order of capture. Dividing the plurality of sections into a plurality of sections of the multiple captured images included in each of The first Shake In quantity In contrast By performing statistical processing , the plurality of section Each of in The amount of blur No. 2 Shake a calculating step of calculating the amount; The plurality of section Each of The second Shake The largest quantity The amount of shaking and an evaluation step of evaluating the blur correction effect based on the result. [Effects of the Invention]

[0008] According to the present invention, it is possible to provide an evaluation method for more accurately determining the image stabilization performance of an imaging device. [Brief explanation of the drawings]

[0009] [Figure 1] FIG. 1 is a schematic diagram showing an example of the configuration of an evaluation system according to a first embodiment. [Figure 2] FIG. 2 is a diagram showing an example of the hardware configuration of an image evaluation apparatus. [Figure 3] FIG. 10 is a block diagram showing the operation of a conventional image evaluation device. [Figure 4] 10 is a graph showing the output waveform of a gyro sensor when stationary. [Figure 5] 10 is a graph illustrating a method for evaluating the effect of image stabilization in a conventional example. [Figure 6] 6 is a graph illustrating a method for evaluating the effect of image stabilization in this embodiment. [Figure 7] 2 is a block diagram showing the operation of the image evaluation device according to the embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0010] Hereinafter, embodiments of the present invention will be described with reference to the drawings, etc. However, the following embodiments do not limit the invention according to the claims, and not all of the features described in the following embodiments are necessarily essential to the present invention.

[0011] FIG. 1 is a schematic diagram showing an example of the configuration of an evaluation system 10 according to this embodiment. The evaluation system 10 is a system for measuring the image stabilization effect of an imaging device 11. The imaging device 11, which is the device under test, is fixed to a vibration table 12. The vibration table 12 mechanically generates vibrations that mimic human hand shake based on input vibration waveform data 13, and vibrates the imaging device 11. The vibration table 12 applies vibrations in the pitch direction (around an axis perpendicular to the plane of the paper in FIG. 1) and the yaw direction (around an axis in the up-down direction in FIG. 1) indicated by arrow 12a. The vibration table 12 can also be controlled to switch between a vibration state and a stationary state.

[0012] An imaging device 11 fixed to a vibration table 12 captures an image of a chart 14 placed directly opposite the imaging device 11. As an example, the chart 14 includes a plurality of black and white stripes having a certain width in the horizontal and vertical directions, and a color natural image portion.

[0013] The image of the chart captured by the imaging device 11 during vibration application is input to the image evaluation device 15. The image evaluation device 15 is, for example, a computer that executes image analysis software. The image evaluation device 15 detects the contrast of the image of the chart captured by the imaging device 11 and measures the deterioration of the image due to vibration application.

[0014] 2 is a diagram showing an example of the hardware configuration of the image evaluation device 15. The image evaluation device 15 has a CPU 101, a ROM 102, a RAM 103, a storage unit 104, an operation I / F 105, a display I / F 106, and an external I / F 107. The elements of the image evaluation device 15 are connected to each other via a system bus 108. Note that CPU stands for Central Processing Unit, ROM stands for Read Only Memory, and RAM stands for Random Access Memory.

[0015] The CPU 101 starts up the OS (Operating System) by a boot program stored in the ROM 102. The CPU 101 executes various processes, which will be described later, by running an image analysis software program stored in the storage unit 104 on the OS. The RAM 103 is used as a temporary area such as the main memory or work area of ​​the CPU 101. In this embodiment, the CPU 101 functions as an acquisition unit that acquires the amount of shake (first change amount), a calculation unit that calculates the amount of shake (second change amount) that has been subjected to statistical processing, a determination unit that determines a judgment value (third change amount), and an evaluation unit that evaluates the effect of shake correction.

[0016] The operation I / F 105 is an interface that connects the CPU 101 and the operation unit 109, and sends information input from the operation unit 109 to the CPU 101. The operation unit 109 is a device that accepts operator input to the image evaluation device 15, and is composed of, for example, a keyboard, a pointing device, etc.

[0017] The display I / F 106 is an interface that connects the CPU 101 and the display unit 110, and outputs image data to be displayed to the display unit 110. The display unit 110 is a device that outputs an operation screen of the image evaluation device 15, and is configured by a display device such as a liquid crystal display.

[0018] The external I / F 107 is an interface for acquiring image data and various information of the image capturing device 11, which is the device under test. The external I / F 107 may be configured to acquire information through wired or wireless communication, or may be configured to read information via a removable storage medium.

[0019] The operation of the image evaluation device 15 in the conventional example will be described using Fig. 3. Fig. 3 is a block diagram showing the operation of the image evaluation device 15 in the conventional example. The output 201 is object image data captured by the imaging device 11 under imaging conditions determined by a plurality of exposure times or brightness when the vibration table 12 is in a vibrating state. The output 201 is processed by the calculation unit 202 in Fig. 3.

[0020] The calculation unit 202 calculates a first change amount based on the output 201. Note that, for simplicity of the following explanation, the first change amount calculated here will be described as the amount of blur. Thus, in this embodiment, the amount of blur is an example of the first change amount. The blur amount data 203 calculated by the calculation unit 202 is calculated according to the number of outputs of the output 201. For example, if the output 201 of the captured subject images is 200, then there will also be 200 pieces of blur amount data 203 to be output.

[0021] The processing unit 204 performs statistical processing on the blur amount data 203. In this embodiment, the processing unit 204 performs statistical processing such as averaging on the blur amount data 203. Thereafter, the evaluation unit 205 determines whether or not the imaging device 11 satisfies the shake correction performance based on the data after the statistical processing.

[0022] 4 and 5, problems with conventional methods for evaluating image stabilization performance will be described. Depending on the components constituting the image stabilization mechanism mounted in imaging device 11 and the control method for the image stabilization mechanism, image stabilization performance may not function stably. As an example, this may be caused by a characteristic of an inertial sensor such as a gyro sensor mounted in the image stabilization mechanism, in which the output of the DC component fluctuates, known as a drift characteristic.

[0023] 4(A) and 4(B) are graphs showing the output waveform of a gyro sensor when the device is stationary. FIG. 4(A) shows the performance of a gyro sensor with extremely low drift characteristics. As shown in FIG. 4(A), in a gyro sensor with extremely low drift characteristics, the output waveform of the gyro sensor when the device is stationary exhibits little fluctuation over time. The gyro sensor in FIG. 4(A) detects mechanical vibrations, human hand shake, and the like as input signals, and outputs that information. The shake correction mechanism installed in the imaging device 11 is driven based on the output from the gyro sensor and performs an operation to suppress shake.

[0024] Figure 4(B) shows the performance of a gyro sensor with relatively large drift characteristics. In Figure 4(B), timing t1 is the timing when power is applied to activate the gyro sensor. Timing t2 indicates the timing a certain period has elapsed since timing t1. Period T1 indicates the range from timing t1 to timing t2.

[0025] It can be seen that the output of the gyro sensor in Figure 4(B) fluctuates significantly during the fixed period T1 when power is applied, even when the sensor is stationary. The output information of the gyro sensor in Figure 4(B), which has the drift characteristics superimposed, contains information other than the actual input. If the image stabilization mechanism performs correction based on such erroneous information, an image with a large amount of blur will be generated.

[0026] FIG. 5 is a graph illustrating a method for evaluating the effect of blur correction in a conventional example. Specifically, FIG. 5 shows the results when blur correction performance is evaluated using the gyro sensor with large drift characteristics shown in FIG. 4(B) and blur amount data is obtained. The evaluation of the effect of blur correction in the conventional example is performed by using the gyro sensor in FIG. 4(B) to calculate the amount of blur using the calculation unit 202 in FIG. 3 based on the object image information obtained by the measurement system in FIG. 1. Note that there is only one exposure condition. In the graph in FIG. 5, the vertical axis represents the amount of blur of the object image. The horizontal axis represents the shooting number, which indicates the amount of blur for each captured image in the order in which they were captured.

[0027] In FIG. 5, the amount of blur at shooting number num_S indicates the amount of blur immediately after shooting begins. Also, the amount of blur at shooting number num_Add indicates the amount of blur at the time of shooting a certain period of time after shooting number num_S. The amount of blur at shooting number num_E indicates the amount of blur at the end of shooting. The amount of blur at interval Data_S is the amount of blur in the divided interval from shooting number num_S to shooting number num_Add. The amount of blur at interval Data_E is the amount of blur in the divided interval from shooting number num_Add to shooting number num_E. Note that a divided interval refers to the interval from a certain shooting number to a subsequent shooting number.

[0028] In the example shown in FIG. 5, the amount of shake does not fluctuate significantly in section Data_E and is stable. On the other hand, it can be seen that the amount of shake fluctuates significantly in section Data_S. These are caused by the characteristics of the gyro sensor shown in FIG. 4(B). Next, conventional evaluation of shake correction performance will be explained in more detail.

[0029] In FIG. 5, the judgment threshold Th for evaluating shake correction performance is indicated by a dot-dash line. The average value Ave calculated by averaging all of the shake amount data is indicated by a dashed line. In conventional methods for evaluating shake correction performance, if the average value Ave is below the judgment threshold Thre, the shake correction performance standard is judged to be met. In other words, if only the comparison result between the judgment threshold Thre and the average value Ave is used, even a shake amount result like that of FIG. 5 will ultimately result in the shake correction performance meeting the evaluation standard. The reason for this is that even if there is a specific shake amount such as that of section Data_S, where the shake amount greatly exceeds the evaluation standard, this tendency will be hidden by calculating the average value of all shake amounts.

[0030] However, with this type of conventional method for evaluating image stabilization performance, it takes a certain period of time for the photographer to realize the evaluated image stabilization performance. In other words, the conventional method for evaluating image stabilization performance has the problem of being unable to evaluate the image stabilization performance while taking into account the stability of the image stabilization.

[0031] The present invention provides an evaluation method for more accurately determining the image stabilization performance of an imaging device by performing an evaluation that takes into account the stability of the image stabilization performance.

[0032] A method for evaluating image stabilization performance in this embodiment will now be described. Fig. 6 is a graph illustrating a method for evaluating image stabilization effect in this embodiment. Note that Fig. 6 is based on Fig. 5, and therefore overlapping parts will be described in the same manner and will not be described again.

[0033] The section Data_ALL shown in Fig. 6 indicates the data length of all data on the amount of shake from shooting number num_S to shooting number num_E. In Fig. 6, sections Data_A, Data_B, and Data_C define the data lengths of the entire shake amount section Dara_ALL when it is divided into three sections.

[0034] Section Data_A indicates the section from shooting number num_S to shooting number num_A. Section Data_B indicates the section from shooting number num_A to shooting number num_B. Section Data_C is the section from shooting number num_B to shooting number num_E. In this embodiment, the average value of the amount of shake is calculated for each of the three sections above and used as the evaluation value. In this manner, in this embodiment, the average value (evaluation value) of the amount of shake in a predetermined range is an example of the second amount of change. Here, the second amount of change is the amount of shake that has been statistically processed (averaged) by dividing the range into sections every predetermined number of shots, as in the example of this embodiment. More specifically, the average value of the amount of shake in a predetermined range is the sum of the amount of shake in the predetermined range divided by the number of shots in the predetermined range. Furthermore, it is preferable that the number of shots be 200 or more in order to perform accurate blur evaluation.

[0035] In FIG. 6, the average value Ave_A indicates the average value of the amount of blur in the section Data_A. The average value Ave_B indicates the average value of the amount of blur in the section Data_B. The average value Ave_C indicates the average value of the amount of blur in the section Data_C. The relationship between the average values ​​of the amount of blur in each of the above sections is shown in the following equation (1). Ave_A>Ave_B≒Ave_C …(1)

[0036] The reason for the result shown in equation (1) is that when using a gyro sensor like the one shown in Figure 4(B) above, shake correction is performed based on erroneous information about startup drift. As a result, the average value Ave_A is larger than the average values ​​in other ranges (average values ​​Ave_B and Ave_C). In the ranges where average values ​​Ave_B and Ave_C are calculated, the output of the gyro sensor is stable and does not fluctuate, so there is not much difference in the results.

[0037] From this result, since the average value Ave_A is the maximum, the section Data_A, which is the divided section for which the average value Ave_A was calculated, is determined as the evaluation section that serves as the basis for shake evaluation. Then, the average value Ave_A of the shake amount in the evaluation section is used as the judgment value to be used for the final evaluation, and the shake correction performance is evaluated. Thus, in this embodiment, the average value Ave_A of the shake amount in the evaluation section is an example of the third change amount (judgment value). The third change amount is the largest second change amount among the second change amounts in each section. Thus, in this embodiment, the section determined to have the largest value related to the shake amount among the multiple divided sections is selected as the evaluation section, and the shake amount in the evaluation section is subjected to statistical processing such as averaging to calculate the judgment value. The shake correction performance is evaluated using this judgment value.

[0038] FIG. 7 is a block diagram showing the operation of image evaluation device 15 in this embodiment. The operation of image evaluation device 15 in this embodiment will be described using FIG. 7. Note that FIG. 7 is based on FIG. 3, and therefore the same descriptions are used for overlapping parts, and the description will be omitted as appropriate. In FIG. 7 as well, as described above, output 201, which is image data of the object captured when vibration is applied, is processed by calculation unit 202 to obtain blur amount data 203.

[0039] The dividing unit 601 divides the blur amount data 203 for the number of shots calculated by the calculation unit 202 into a plurality of divided sections.

[0040] The calculation unit B602 performs statistical processing such as averaging for each of the multiple sections divided by the division unit 601, and calculates an evaluation value of the image stabilization performance.

[0041] The comparison unit 603 selects the maximum evaluation value from the evaluation values ​​of the image stabilization performance calculated by the calculation unit B 602. The comparison unit 603 transmits the divided section for which the maximum evaluation value has been calculated as the evaluation section to the processing unit 604.

[0042] The processing unit 604 uses the information on the evaluation interval obtained by the comparison unit 603 to calculate an average blur amount value, which is a judgment value, for the data corresponding to the evaluation interval out of the blur amount data 203 .

[0043] The evaluation unit 205 determines whether or not the shake compensation performance is satisfied based on the average shake amount value obtained by the processing unit 604. As described above, the determination of whether or not the shake compensation performance is satisfied is made based on whether or not the determination threshold Thre is exceeded. That is, if the determination value is equal to or greater than the determination threshold Thre, it is determined that the shake compensation performance is not satisfied. On the other hand, if the determination value is less than the determination threshold Thre, it is determined that the shake compensation performance is satisfied.

[0044] When no division into sections is performed, the blur amount data 203 obtained by the calculation unit 202 is transmitted directly to the processing unit 604 without going through the division unit 601 .

[0045] Although the preferred embodiments of the present invention have been described above, the present invention is not limited to these embodiments and various modifications and changes are possible within the scope of the gist of the present invention.

[0046] In this embodiment, for simplicity of explanation, the judgment value is described as the average value of the amount of shake. However, the judgment value is not limited to the average value of the amount of shake. For example, the variance value of the amount of shake may be used as the judgment value. Even when the variance value is used, a predetermined variance value is used as a judgment threshold, and if the variance value is equal to or greater than the judgment threshold, it is determined that the shake compensation performance is not satisfied. On the other hand, if the variance value is less than the judgment threshold, it is determined that the shake compensation performance is satisfied.

[0047] Similarly, in this embodiment, the statistical processing for calculating the evaluation value is an averaging process that finds the average value for each divided section, but this is not limited to this. For example, the same effect can be obtained by performing a process that finds the variance value for each divided section.

[0048] Furthermore, in the calculation of the evaluation value in this embodiment, three divided sections are set, but the number of sections is not limited to three. That is, when a predetermined number of captured images (200 or more in this embodiment) is divided into multiple sections, three sections are used in this embodiment. Specifically, the multiple sections include a divided section (first divided section) including the first image at the start of capture, a divided section (second divided section) that follows the first divided section over time, and a divided section (third divided section) including the last image at the end of capture. However, the second divided section is not required, and there may be two divided sections overall. Furthermore, by dividing the second divided section into multiple sections, there may be four or more divided sections overall.

[0049] When dividing sections as in this embodiment, it is desirable that each divided section contains approximately 70 captured images, which is the number of pieces of data. If the number of pieces of data is too small when dividing sections, it is not possible to eliminate the effects of singular points in the acquired data, which may result in a decrease in the accuracy of evaluating image stabilization performance. Therefore, it is necessary to divide sections so that the number of captured images, which is the number of pieces of data, is approximately 70. For example, if 500 pieces of data are acquired in total, the number of divided sections should be approximately 7. If 700 pieces of data are acquired in total, the number of divided sections could be 10.

[0050] The method of this embodiment makes it possible to perform an evaluation with higher accuracy than conventional methods, even in cases where performance is unstable, such as when the amount of shake tends to fluctuate greatly in the early stages due to factors such as the components and controls that make up the shake correction, without the unstable results being buried in the average.

[0051] The present invention can also be realized by supplying a program that realizes one or more functions of the embodiments to a system or device via a network or a storage medium, and having one or more processors in the computer of the system or device read and execute the program. The present invention can also be realized by a circuit (e.g., ASIC) that realizes one or more functions. [Explanation of symbols]

[0052] 11 Imaging device 12 Shaking table 13 Excitation waveform data 14 Charts 15 Image evaluation device 202 Arithmetic section 203 Shake amount data 205 Evaluation Department 601 Split section 602 Arithmetic unit B 603 Comparison Section 604 Processing section

Claims

1. A method for evaluating a shake correction effect of an imaging device, comprising: an acquisition step of acquiring a first amount of blur for each of a predetermined number of subject images captured by the vibrated imaging device; a calculation step of dividing the predetermined number of captured subject images into a plurality of sections such that each section includes a plurality of captured images in the order in which they were captured, and performing statistical processing on the first amount of blur of the plurality of captured images included in each of the plurality of sections to calculate a second amount of blur that is the amount of blur in each of the plurality of sections; an evaluation step of evaluating a shake correction effect based on the largest of the second shake amounts in each of the plurality of sections. An evaluation method characterized by:

2. The predetermined number of shots is 200 or more. The evaluation method according to claim 1 .

3. The statistical processing is a process for calculating an average value.

3. The evaluation method according to claim 1 or 2.

4. The statistical processing is a process for determining a variance value.

3. The evaluation method according to claim 1 or 2.

5. When the predetermined number of shots is divided into a plurality of consecutive sections, the plurality of sections include a first divided section including the first shot at the start of shooting to the nth shot, a second divided section including the n+1th shot to the mth shot following the first divided section, and a third divided section including the i-th shot, which is a number greater than m, to the last shot at the end of shooting. The evaluation method according to any one of claims 1 to 4.

6. In the evaluation step, if the largest of the shake amounts that have been subjected to the statistical processing is equal to or larger than a predetermined threshold, it is determined that the shake compensation performance is not satisfied, and if the largest of the shake amounts that have been subjected to the statistical processing is smaller than the predetermined threshold, it is determined that the shake compensation performance is satisfied. The evaluation method according to any one of claims 1 to 5.

7. An apparatus for evaluating the effect of image stabilization of an imaging device, comprising: an acquisition means for acquiring a first amount of shake for each of a predetermined number of subject images captured by the vibrated imaging device; a calculation means for dividing the predetermined number of captured subject images into a plurality of sections such that each section includes a plurality of captured images in the order in which they were captured, and for calculating a second amount of blur, which is the amount of blur in each of the plurality of sections, by performing statistical processing on the first amount of blur of the plurality of captured images included in each of the plurality of sections; and an evaluation unit that evaluates a shake correction effect based on the largest of the second shake amounts in each of the plurality of sections. An evaluation device characterized by:

8. A program for causing a computer to evaluate the image stabilization effect of an imaging device, an acquisition step of acquiring a first amount of blur for each of a predetermined number of subject images captured by the vibrated imaging device; a calculation step of dividing the predetermined number of captured subject images into a plurality of sections such that each section includes a plurality of captured images in the order in which they were captured, and performing statistical processing on the first amount of blur of the plurality of captured images included in each of the plurality of sections to calculate a second amount of blur that is the amount of blur in each of the plurality of sections; an evaluation step of evaluating a shake correction effect based on the largest of the second shake amounts in each of the plurality of sections. A program characterized by:

Images