JP4103129B2 - Solid-state imaging device and imaging method - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an imaging method for generating an image signal with an extended effective dynamic range, and a solid-state imaging device such as a video camera using the method.
[0002]
[Prior art]
The imaging device is configured by a solid-state imaging device such as an imaging tube or a CCD (Charge Coupled Device), and is widely used in imaging units such as a digital camera, a scanner, or a video camera. Such an imaging device has a narrow dynamic range compared to the conventional silver halide photography system, and in bright areas of the subject, the brightness level is extremely high when the subject is backlit. In the dark part, the brightness level is extremely low, and the image may become too black (blackout). Therefore, conventionally, in order to eliminate such a phenomenon and prevent deterioration in image quality, a method of synthesizing one image from a plurality of images with different exposure amounts obtained from the same subject has been adopted.
[0003]
For example, Japanese Patent Application Laid-Open No. 2-174470 and Japanese Patent Application Laid-Open No. 5-7336 describe a technique for correcting image shift due to camera shake or the like, which has been a problem when performing image composition. The image composition method employed therein is a general one. According to the above-mentioned Japanese Patent Laid-Open No. 5-7336, as shown in FIG. 11, a signal from the image sensor is output with two types of exposure time of 1/1000 second and 1/60 second in one frame period. Are respectively assigned to the odd (ODD) field and the even (EVEN) field, which are images taken with an exposure time of 1/1000 second and 1/60 second. Since these images have different exposures, even if “overexposure” or “blackout” occurs in one field, it does not occur in the other field. In this way, the dynamic range is expanded and the image quality is improved by interchanging the usable portions of the two images.
[0004]
[Problems to be solved by the invention]
However, in the conventional method, since each image is taken in order, there are two major problems. The first problem is that since only about two images with different exposure times can be obtained, the composition is difficult. Of course, in principle, two or more images can be combined, but as the number of images increases, the entire imaging period becomes longer. Therefore, in order to suppress temporal changes between images, to keep the total imaging period within one field or one frame, and to make the combined image signal comply with standards such as the NTSC system, two or more images are required. It is difficult to do. Therefore, in practice, two images are used for synthesis. In this case, the exposure time is selected to be relatively different in order to extend the effective dynamic range of the composite image. In that case, it is necessary to adjust the characteristics of the mutual image signals with respect to the amount of incident light in order to prevent an unnatural contrast change at the boundary between the two. That is, gain adjustment is performed by amplifying the characteristic with the shorter exposure time, but this time, there is a problem that noise frequently occurs at the boundary between images.
[0005]
The second problem is that images with different exposure times are taken in different periods. For this reason, when there is a movement in the imaging target, information is lost at the boundary portion of the target during synthesis. The technique described above eliminates the positional shift between images over time by performing coordinate conversion that shifts the images according to the main subject. However, the shifted peripheral portion of each image is discarded, and the area common to the two images is enlarged to the original image size.
[0006]
As described above, conventionally, some measures must be taken for contrast and positional deviation at the time of composition, and as a result, the image to be captured is not faithfully reproduced. It was.
[0007]
The present invention has been made in view of such a problem, and an object thereof is to provide a solid-state imaging device and an imaging method capable of easily obtaining an image with a wide effective dynamic range and suppressing image shift and noise.
[0008]
[Means for Solving the Problems]
The solid-state imaging device according to the present invention continuously captures a plurality of times within a predetermined time, and adds an imaging unit that generates a plurality of unit images that are sequentially captured, and the unit images. A composite image generating unit that generates a plurality of composite images that overlap and have different exposure times, and a composite unit that combines the composite images to create a composite image with an expanded dynamic range. Is. Further, the composition image generation means generates all the composition images so that there is a common overlapping portion in the imaging period, and the common overlapping portion includes the unit image captured last. It has become.
[0009]
The imaging method of the present invention continuously performs imaging a plurality of times within a predetermined time, generates a plurality of unit images that are sequentially captured, adds the unit images, and partially overlaps the imaging periods; and A plurality of compositing images having different exposure times are generated, and the compositing images are further combined to create a composite image with an expanded dynamic range. Further, all of the images for synthesis are generated so as to have a common overlapping portion in the imaging period, and this common overlapping portion includes the unit image captured last.
[0010]
In the solid-state imaging device and imaging method of the present invention, unit images are added in consideration of the number and position with time, and a plurality of composite images having different exposure times are generated so as to partially overlap imaging periods. Is done. Further, all of the synthesis images are generated so as to have a common overlapping portion in the imaging period, and this common overlapping portion includes the unit image captured last.
[0011]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.
[0012]
FIG. 1 is a block diagram showing a configuration of a solid-state imaging apparatus according to an embodiment of the present invention. FIG. 2 conceptually shows a state in which a compositing image is generated from a unit image, and FIG. 3 is a time chart showing a relationship between an imaging period and an exposure time between the unit image and the compositing image. In this solid-state imaging device, a plurality of compositing images 22 whose imaging periods partially overlap and whose exposure times are different from each other are generated from the unit image 21 actually captured and further used. A composite image 23 is generated. Note that “imaging period” refers to the period during which images are captured on the time axis, and “exposure time” refers to the length of time during which exposure is performed and images are captured. And
[0013]
The optical system 11 forms an image of incident light on the image sensor 12, and the image sensor 12 is formed of a solid-state image sensor such as a CMOS sensor or a CCD sensor. _This is converted to _V and output to the camera signal pre-stage processing circuit 13.
[0014]
The imaging element 12 is controlled by the drive circuit 16 so that imaging is continuously performed a plurality of times (n times) within a predetermined time and n unit images 21 are sequentially captured (FIG. 3). Imaging conditions at that time, that is, the length of the entire imaging period (total exposure time), the number of unit images, and the respective exposure times are set according to the relationship between each exposure time and the number of images of the composition image 22 described later. .
[0015]
For example, a series of imaging is performed within one frame or one field period, and further, within a predetermined period. Here, all the unit images 21 are captured with a predetermined unit exposure time t. The unit exposure time t can be set as appropriate. For example, the unit exposure time t may be approximately equal to or less than the exposure time (1/1000 second) of the conventional image for synthesis. For example, the length of the imaging period can be set to 1/60 seconds, the unit exposure time 1/6000 seconds, and 10 unit images 21. Such high-speed imaging can be sufficiently realized at present for some CMOS sensors.
[0016]
The camera signal pre-stage processing circuit (pre-stage processing circuit) 13 has a function of converting the voltage signal S _V corresponding to each unit image 21 from the image sensor 12 into image data Du corresponding to n unit images 21. ing. That is, in the pre-processing circuit 13, with respect to the voltage signal S _V, well known process similar to the signal processing in the conventional video camera including an A / D conversion is performed. The image data Du of the unit image 21 generated here is output to the image memory 14.
[0017]
The image memory 14 stores the unit image data Du input from the pre-stage processing circuit 13 and outputs the unit image data Du in response to a read / write request from the camera signal post-stage processing circuit (post-stage processing circuit) 15. And a function of storing the composition image data Ds input from the post-processing circuit 15. In addition, a work area associated with image composition is provided for the post-processing circuit 15. The image memory 14 is a device that can simultaneously read and write a plurality of data, and is specifically configured as a storage device such as a hard disk or a semiconductor memory.
[0018]
The post-processing circuit 15 has a function of adding n unit images 21 in the order of imaging, and generating a plurality of compositing images 22 having overlapping imaging periods and different exposure times. . That is, by adding m pieces of continuous unit images 21 (m = 1 to n), the exposure time is m times the unit exposure time t, and the imaging period is the sum of the added imaging periods of the unit images 21. The synthesis image 22 is obtained. This operation reads image data Du of n unit images 21 from the image memory 14, adds these image data Du in units of pixels according to the exposure time and imaging period, and image data Ds corresponding to the composition image 22. This is done by generating The obtained image data Ds is stored in the image memory 14.
[0019]
Various methods of selecting the unit image 21 used for generating the composition image 22 are conceivable. As a simple example, as shown in FIG. 2 and FIG. 3, 1 image, 2 images,..., N images from the last image are added, respectively, to obtain n composition images 22. A method is mentioned. In this case, the composition image 22 shares at least an imaging period corresponding to the last unit image 21, and the exposure times are t, 2t,. It is only necessary to generate any number of such composition images 22 necessary for composition of the images, and the exposure time and imaging period of each composition image 22 can be appropriately set as necessary. However, in order to finally obtain a smoothly synthesized image, it is desirable to obtain three or more images for synthesis 22 here. In addition, in order to suppress the positional deviation of the subject at the time of composition, it is desirable to set the imaging period of the composition image 22 so that the overlapping portions are as large as possible.
[0020]
FIG. 4 shows the characteristics of the image data Ds with respect to the amount of incident light. As described above, the composition image 22 (hereinafter abbreviated as image 22) has different characteristics of the obtained signal value, that is, the value of the image data Ds, depending on the exposure time. The characteristic 42 in the case of the exposure time 2t has a slope twice that of the data characteristic 43 of the image 22 that is the exposure time t, and the slope of the characteristic 41 in the case of the exposure time nt has a slope n times. That is, when the amount of incident light is small and the subject is dark, if the exposure time is lengthened as in the characteristic 41 and the obtained data signal value is sufficiently large, the gradation can be maintained even when it is dark. On the other hand, for a bright subject with a large amount of incident light, if the exposure time is shortened as in characteristic 43, the data signal does not reach the saturation point even with a large amount of incident light, so that the gradation is maintained even if it is bright.
[0021]
As described above, since the exposure time, that is, the data characteristics of the images 22 are different from each other, an appropriate gradation (gradient of characteristics) can be given to a wide range of incident light amounts as a whole. All the values of the image data Ds are saturated at a certain upper limit value.
[0022]
Further, the post-processing circuit 15 has a function of performing synthesis using these images 22 and creating a synthesized image 23 with an expanded effective dynamic range. Any combination method may be used, but as commonly used, one signal value is amplified at the boundary between images having discontinuous characteristics so as to eliminate the joint of the images. In this method, if this operation is not successful, an unnatural region may occur due to a relative change in the magnitude of signal values of images to be joined. Therefore, in the present embodiment, the following method is adopted in order to avoid such problems.
[0023]
In other words, the post-processing circuit 15 selects an image data Ds suitable for the incident light amount from the image data Ds of each image 22 and uses it for synthesis. First, in the image 22 having the longest exposure time, if the image data Ds is smaller than the set upper limit value, it is adopted as it is. Here, the set upper limit value is, for example, the value of the saturation point of the image data Ds, but it may be set to a value lower than the saturation point to guarantee the image quality. The same operation is performed using the image 22 with the exposure time (n−1) t for the pixels for which the image 22 with the exposure time nt has not been adopted. The same operation is repeated until the image data Ds to be adopted for all the pixels is determined. For the last remaining pixel, the image data Ds of the image 22 at the exposure time t is employed.
[0024]
At this time, in the present embodiment, the value of the image data Ds of each image 22 is corrected so that the data characteristic with respect to the incident light amount becomes one continuous characteristic. Although details will be described later, the image data Ds of the image 22 having the longest exposure time (m is an integer of 2 or more) in the image 22 has an exposure time of No. 1 in its own displacement from the signal lower limit value in the same image. All the displacements from the lower limit value to the set upper limit value of the image data Ds in each of the images 22 from the longest to the (m-1) th longest are converted into integrated values and then combined (see FIG. 6). As described above, the magnitudes of the data signal values of the respective images 22 are relativized and joined together without breaks, so that the dynamic range is appropriately expanded.
[0025]
Note that the post-processing circuit 15 performs the above-described image synthesis by sequentially superimposing the images 22, and the processing to be performed on the image data Ds of each image 22 is known in advance. The image data Ds may be processed in parallel, and the processed image data Dw may be applied to the pixel position of the original image data Ds to create a single composite image 23.
[0026]
In addition, the post-processing circuit 15 performs other known processes similar to gamma correction and signal processing of a normal video camera. In the present embodiment, the post-processing circuit 15 corresponds to “compositing image generating means” and “combining means” of the present invention.
[0027]
The drive circuit 16 is a circuit for driving the image sensor 12. The AE control unit 17 has a function of performing aperture adjustment of the optical system 11, electronic shutter control of the drive circuit 16, gain adjustment of the pre-processing circuit 13, and the like based on exposure state information input from the post-processing circuit 15. The exposure conditions for the images 21 to 23 are adjusted.
[0028]
Next, the operation of the solid-state imaging device and the imaging method according to the present embodiment will be described.
[0029]
In the following, as a more specific example, an imaging period corresponding to the last unit image 21L is shared from eight unit images 21 taken at unit exposure time t as shown in FIG. When four images 22A, 22B, 22C, and 22D that are t, 2t, 4t, and 8t, that is, 2 ⁿ times the unit exposure time t (n is an integer of 0 or more), are used for image synthesis. Will be explained.
[0030]
First, the optical system 11 and the imaging device 12 sequentially capture eight unit images 21 with a unit exposure time t. The image sensor 12 outputs a voltage signal S _V corresponding to each unit image 21 to the pre-processing circuit 13. The pre-stage processing circuit 13 converts the input voltage signal S _V into image data Du corresponding to eight unit images 21 and stores the image data Du in the image memory 14.
[0031]
Next, the post-processing circuit 15 reads the image data Du of the eight unit images 21 from the image memory 14, and adds the image data Du in units of pixels, thereby sharing the imaging period corresponding to the last unit image 21L. At the same time, image data Ds corresponding to four images 22A, 22B, 22C, and 22D having exposure times t, 2t, 4t, and 8t, respectively, is generated (FIG. 5). Since these images 22A to 22D share the imaging period of at least one unit image 21 (last unit image 21L), it is difficult for the subject to be displaced with time during that period, and to avoid missing images. Is done. The obtained image data Ds is stored in the image memory 14.
[0032]
Further, the post-processing circuit 15 combines the images 22A to 22D to create a combined image 23. FIG. 6 is a characteristic diagram with respect to the incident light amount of the data signal in accordance with this specific example. FIG. 6A corresponds to before synthesis and FIG. 6B corresponds to after synthesis. Note that each characteristic in FIG. 5A is distinguished by the code at the end of the corresponding images 22A to 22D.
[0033]
First, in the image 22D having the longest exposure time, if the image data Ds is a value smaller than the saturation point value m of the image data Ds, it is directly adopted as the combined image data Dw. Therefore, the range from 0 to m in the image data Dw is constituted by the image data Ds of the image 22D having the longest exposure time. This corresponds to the incident light amount range LD.
[0034]
For the pixels for which the image data Ds of the image 22D has not been adopted, the image 22C having the exposure time 4t is referred to, and if the image data Ds is smaller than the saturation value m, it is used as the image data Dw. . Therefore, what is used in the image data Ds of the image 22C covers the range LC of the incident light quantity that is saturated in the image 22D but not saturated in the image 22C. Here, since the gradient of the characteristics of the images 22D and 22C is 2: 1, the range of the image data Ds used is from m / 2 to m.
[0035]
When the image data Ds corresponding to the range LC is synthesized, the maximum value m of the image data Dw in the range LD is added to its own displacement value (0 to m / 2) from the lower limit m / 2 in the image 22C. The converted value is converted into image data Dw. As a result, as shown in FIG. 6B, the combined image data Dw has a characteristic in which two different characteristics D and C are continuously combined. For the image 22C, the low-frequency image data Ds less than m / 2 is not used, so that the noise of the image data Dw is reduced.
[0036]
Hereinafter, the same operation is performed by sequentially referring to the images 22B and 22A until there are no pixels for which the image data Ds to be used is not determined. In the image 22B, a part of m / 2 to m corresponding to the incident light amount range LB is used in the image data Ds. At the time of synthesis, the image 22B is converted to a value obtained by adding the maximum value 3 m / 2 of the image data Dw up to the range LC to the displacement value (0 to m / 2) from the lower limit m / 2 in the image 22B. Let it be data Dw.
[0037]
In the image 22A, a part of m / 2 to m corresponding to the range LA of the incident light quantity is used in the image data Ds. At the time of synthesis, the image data Dw is converted into a value obtained by adding the maximum value 2m of the image data Dw up to the range LB to the displacement value (0 to m / 2) from the lower limit m / 2 in the image 22A. And
[0038]
In this way, composition is performed, and a composite image 23 having the data signal characteristics shown in FIG. 6B is created. In this composition method, since the processing to be performed on the image data Ds of each of the images 22A to 22D is known in advance, the image data in the usage range (m / 2 to m) is actually applied to each of the images 22A to 22D. A single composite image 23 may be created by performing the processing of Ds in parallel and applying the processed image data Dw to the pixels.
[0039]
In the composite image 23, the data characteristics with respect to the incident light amount of each of the images 22A to 22D are continuously connected, and the effective dynamic range is expanded. For the images 22A to 22C excluding the image 22D, the low-frequency image data Ds less than m / 2 is not used, so that the noise of the image data Dw is reduced as a whole. The composite image 23 is output as image data Dw from the post-processing circuit 15.
[0040]
As described above, in the present embodiment, since the compositing images 22 having different exposure times are generated from the unit images 21, three or more compositing images 22 more than usual can be easily obtained. In the composite image 23 obtained from the plurality of composite images 22, a smoother halftone can be expressed. In addition, since a series of images for synthesis 22 is generated so as to share at least one unit image 21, the temporal change between the images for synthesis 22 can be reduced, and image loss due to subject position deviation or the like can be reduced. It can be avoided.
[0041]
Further, the number of images for synthesis 22, the exposure time, and the imaging period can be freely changed. Each exposure time of the composition image 22 can be given a certain correlation such that it can be expressed as a function based on the setting of the exposure time of the unit image 21. Therefore, the exposure time of the plurality of composition images 22 can be given. (Characteristics) can be selected in association with each other. Further, such an operation determines, for example, a range used for the synthesis of the image data Ds. Here, since the exposure time of the composition image 22 is 2 ⁿ times the unit exposure time t, all usable data values of each composition image 22 may fall within a certain range of m / 2 to m. it can. Therefore, since the low frequency component (0 to m / 2) of each synthesis image 22 is not used at the time of synthesis, noise in the image 23 can be reduced. In addition, by using 2 ⁿ times, handling is easy when image data processing is performed as bit data.
[0042]
In the present embodiment, the exposure time of the compositing image 22 is set to n times or 2 ⁿ times the unit exposure time t (n is an integer of 0 or more), but the relationship between the compositing images is another function. It may be represented by For example, the exposure time of each composite image is set to n ^k times such as 3 ^k times (1, 3, 9,...) Of the unit exposure time t (n is a fixed value, an integer of 2 or more, k is a variable) And an integer greater than or equal to 0). In this case, as in the case of 2 ⁿ times, since the application range (upper limit value and lower limit value) for synthesis matches in all the synthesis images, it is easy to handle as a synthesis method.
[0043]
Regarding the setting of the imaging period, here, the imaging period shared by the compositing image 22 is set to be the imaging period of the unit image 21 captured last. However, the compositing image of the present invention has its imaging period. For example, as shown in FIG. 7, the sharing period may be set at the beginning of the entire imaging period. Further, as shown in FIG. 8, the sharing period may be set at the center of the entire imaging period. In this case, it is not necessary to consider the time difference between the images for synthesis, and the effect of reducing the positional shift of the imaging target with time is high. Furthermore, as illustrated in FIG. 9, the imaging periods may partially overlap with one of the compositing images.
[0044]
【Example】
Next, specific examples of the present invention will be described. In the following examples, the configuration and operation of the entire solid-state imaging device are basically the same as those in the embodiment.
[0045]
FIG. 10 is a characteristic diagram with respect to the amount of incident light of a data signal according to the embodiment of the present invention, where (A) is before synthesis and (B) is after synthesis. In this embodiment, unit images are sequentially captured at unit exposure time t, and from these unit images, an imaging period corresponding to at least one unit image is shared, and exposure times are t, 2t, 4t, 8t, and 16t, respectively. 5 composite images were generated.
[0046]
Next, these images for synthesis are synthesized. If the image data is used as it is as in the embodiment, the data length after synthesis is particularly large, and a large area is required in the image memory. . Therefore, the image data of each composition image is reduced in advance at a certain ratio so that the synthesized data signal characteristics can be obtained in a normalized form as shown in FIG. went. The solid line portion in FIG. 7B is the data signal characteristic after synthesis. From this figure, it can be confirmed that the dynamic range is expanded 16 times.
[0047]
In addition, since the characteristics of the image data after synthesis is a polygonal line, a process for correcting to a logarithmic curve is effective. Here, the corrected curve is indicated by a dotted line in FIG. This curve is
(Data signal value) = log {50 × (incident light quantity) +1}} / log 51
Is a logarithmic curve represented by As a result, the data signal characteristics after synthesis can be smoothed, and an effect of improving the image quality is expected. Note that it is generally known that humans treat the amount of light entering the eye with logarithmic characteristics and feel it as bright, so logarithmic correction is what makes the composite image look physiological. It can be said that it is close to.
[0048]
In addition, this invention is not limited to the said embodiment, A various deformation | transformation implementation is possible. For example, in the above embodiment, all the unit images 21 are captured with the unit exposure time t, but the exposure time of each unit image can be arbitrarily selected and does not have to be the same value. However, if the exposure time of the unit image is kept constant as in the embodiment, the exposure time of the image for synthesis is proportional to the number of unit images to be added, so that control becomes easier.
[0049]
Further, the unit image capturing conditions and the method of selecting the unit images used for generating the composition image may be determined in advance as fixed conditions, but may be changed according to the condition of the subject.
[0050]
Furthermore, in the above embodiment, the composition image 22 is generated by the data processing in the post-processing circuit 15. However, when the unit image to be added is known in advance, the input signal is converted into an image by the pre-processing circuit. It is also possible to take a method of performing arithmetic processing as it is converted into data, and storing the value in the image memory as a synthesis image when a predetermined unit image is added.
[0051]
【The invention's effect】
As described above, according to the solid-state imaging device of the present invention and the imaging method of the present invention, a plurality of unit images are sequentially captured within a predetermined time, and a plurality of unit images that are sequentially captured are generated. And generating a plurality of compositing images whose imaging periods partially overlap and having different exposure times, and compositing the compositing images to create an image with an extended dynamic range , Since all the images for synthesis are generated so as to have a common overlapping part in the imaging period, and this common overlapping part includes the unit image captured at the end , a plurality of images and the imaging period of each other Can be obtained by a very simple method. Therefore, since the obtained image is synthesized from a plurality of synthesis images, not only the dynamic range is expanded but also a smooth halftone can be expressed.
Further, since a plurality of synthesis images are used, the low-frequency component of the signal of each synthesis image is not used for synthesis, so that noise in the obtained image can be reduced. At the same time, the imaging periods of the compositing images partially overlap each other, so that changes with time of each other are reduced, and it is possible to avoid image loss due to subject position deviation or the like.
Further, such a composition image can be changed in the number of images, exposure time, and imaging period only by changing the number of unit images used for generation and the position over time without changing the imaging conditions such as the exposure time of the unit image. be able to. As described above, the degree of freedom in setting conditions in image synthesis is extremely high, and it is possible to easily set conditions.
[Brief description of the drawings]
FIG. 1 is a block diagram illustrating a configuration of a solid-state imaging device according to an embodiment of the present invention.
FIG. 2 is a conceptual diagram for explaining a procedure for generating a synthesis image in the solid-state imaging device shown in FIG. 1;
3 is a time chart illustrating a relationship between an image for synthesis and a unit image generated in the solid-state imaging device shown in FIG. 1 in an imaging period and an exposure time. FIG.
4 is a characteristic diagram of a data signal of a synthesis image generated by the solid-state imaging device shown in FIG. 1. FIG.
5 is a time chart showing a relationship between a specific example of a composition image generated by the solid-state imaging device shown in FIG. 1 and a unit image in an imaging period and an exposure time. FIG.
6A and 6B are diagrams for explaining a method of synthesizing an image using the image for synthesis shown in FIG. 5, in which FIG. 6A is a characteristic diagram before synthesis, and FIG. 6B is a characteristic diagram after synthesis.
7 is a time chart showing a first modified example of a relationship between an imaging period and an exposure time between the composition image and the unit image shown in FIG. 3;
8 is a time chart showing a second modification of the relationship between the imaging period and the exposure time between the composition image and the unit image shown in FIG. 3;
FIG. 9 is a time chart showing a third modification of the relationship between the imaging period and the exposure time between the composition image and the unit image shown in FIG. 3;
FIGS. 10A and 10B are diagrams for explaining a method for synthesizing an image in the solid-state imaging device according to the embodiment of the present invention. FIG. 10A is a characteristic diagram before synthesis, and FIG. 10B is a characteristic diagram after synthesis.
FIG. 11 is a time chart for explaining an image composition method in a conventional solid-state imaging device;
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 11 ... Optical system, 12 ... Image sensor, 13 ... Camera signal pre-processing circuit, 14 ... Image memory, 15 ... Camera signal post-processing circuit, 16 ... Drive circuit, 17 ... AE control part, 21 ... Unit image, 22, 22A ˜22D... Composition image, 23... Composition image, 41, 42, 43... Data signal characteristics of composition image, Sv... Voltage signal, t... Unit exposure time, Du. Image data, Dw... Composite image image data, LA to LD...

Claims (19)

Imaging means for continuously performing imaging a plurality of times within a predetermined time, and generating a plurality of unit images sequentially captured;
Adding the unit images to generate a plurality of compositing images that have overlapping imaging periods and different exposure times from each other; and
Synthesizing the synthesis image to create a synthesized image with an extended dynamic range , and
The composition image generation means generates all the composition images so that there is a common overlapping part in the imaging period,
The solid-state imaging device, wherein the common overlapping portion includes a unit image captured last . The solid-state imaging device according to claim 1, wherein the composition image generation unit generates three or more images for composition. The solid-state imaging device according to claim 1, wherein all the unit images are captured with the same predetermined exposure time. The exposure time of each of the images for synthesis is ^nk times the predetermined exposure time (n is a fixed value and an integer of 2 or more, k is a variable and an integer of 0 or more). 3. The solid-state imaging device according to 3 . 5. The solid-state imaging device according to claim 4 , wherein an exposure time of each of the synthesis images is 2 ^k times the predetermined exposure time (k is a variable and an integer of 0 or more). 2. The solid-state imaging according to claim 1, wherein the synthesizing unit extracts, for each of the pixels, a signal belonging to an image that is smaller than a set upper limit value and has the longest exposure time from each of the images for synthesis. apparatus. The signal value of the m-th image with the longest exposure time (m is an integer equal to or greater than 2) of the image for synthesis has its own displacement from the signal lower limit in the same image, and the exposure time of the image for synthesis has 7. The system according to claim 6, wherein the displacement from the lower limit value of the signal to the set upper limit value in each of the longest to the (m-1) longest is converted into an integrated value and used for synthesis. Solid-state imaging device. The solid-state imaging device according to claim 6, wherein the set upper limit value of the image signal is a signal value at a saturation point with respect to an incident light amount. The synthesis means reduces the image signal of each of the synthesis images so that the dynamic range of the synthesis image is normalized before synthesizing the synthesis image. 6. The solid-state imaging device according to 6 . The solid-state imaging device according to claim 1, wherein a CCD sensor or a CMOS sensor is used as the imaging unit. An imaging method for correcting image quality of an image obtained by imaging a subject,
Continuously capture multiple times within a given time, generate multiple unit images that are sequentially captured,
Adding the unit images to generate a plurality of compositing images having overlapping imaging periods and different exposure times from each other;
Furthermore, the composite image is combined to create a composite image with an expanded dynamic range ,
All of the images for synthesis are generated so as to have a common overlapping part in the imaging period,
The imaging method according to claim 1, wherein the common overlapping portion includes a unit image captured last . Imaging method of claim 1 1, wherein the generating three or more of the synthesis image. Imaging method of claim 1 1, wherein the imaging the unit images in all the same predetermined exposure time. The exposure time of each of the images for synthesis is ^nk times the predetermined exposure time (n is a fixed value and an integer of 2 or more, k is a variable and an integer of 0 or more). The imaging method according to 13 . The imaging method according to claim 14 , wherein an exposure time of each of the images for synthesis is 2 ^k times the predetermined exposure time (k is a variable and is an integer of 0 or more). From each of the combined images, smaller than the set upper limit value, a signal exposure time belongs to the longest image extracted for each pixel, the imaging method of claim 1 1, wherein a used for the synthesis. The signal value of the m-th image with the longest exposure time (m is an integer of 2 or more) among the images for synthesis is changed to its own displacement from the signal lower limit in the same image, and the exposure time of the image for synthesis is 1 Banchou converts all integrated value of displacement from the lower limit value of the signal at each of the up to long to (m-1) th to the set upper limit value from the casting claim 1 6, wherein the use in the synthesis Imaging method. The set upper limit value of the image signal, the imaging method of claim 1 6, wherein the the signal value at the saturation point with respect to the amount of incident light. Before synthesizing the synthesized image, the imaging of claims 1 to 6, wherein the dynamic range of the composite image, characterized in that to reduce the respective image signals of the combined images such that those normalized Method.

Images