JP5138510B2 - Image data processing device - Google Patents

Description

The present invention may, for example, relates to a digital camera, regarding the images data processing equipment which subjects the image data is Ru can be determined whether taken with backlighting conditions.

  Conventionally, for example, for a camera such as a digital camera, the image data of the preview acquired before the actual shooting is analyzed, and when the subject near the center of the image data of the preview is darker than the background, it is determined that the backlight is in a dark state. That is, a configuration in which a flash is emitted to compensate for the brightness of the subject is known. This backlight determination is for correcting the backlight by irradiating a flash on a dark part, and is a detection function provided in the automatic exposure (AE) of the camera. And since the flash is irradiated toward the center of the screen, the backlight condition is only when the center is dark and the periphery is bright and the subject is in the range where the flash irradiation light reaches. Judgment.

  However, the above backlight determination method is based on the premise of flash irradiation. In a situation where the center of the screen is bright and surrounding buildings are dark, the backlight is not determined and the flash is not emitted. In addition, even if the flash is emitted by manual operation, etc., the flash is irradiated toward the center of the screen, and the reachable distance of the flash irradiation is limited, so the flash irradiation light It rarely reaches the dark part and has a problem that a darkly crushed area is generated in the image data.

  On the other hand, a method is known in which post-shooting image data is analyzed and luminance data is adjusted by post-processing (for example, see Patent Document 1). This method corrects the brightness data in the dark area when the brightness data is small (low brightness) and there is a dark area in the image data, preventing the occurrence of situations such as blackening of the subject. I am trying.

However, for example, when the subject person is wearing black clothes, even if he wants to maintain the color of the black clothes, the process of increasing the luminance data is performed more than necessary, and a specific color or noise may appear strongly. Have a problem. In addition, for example, a checkerboard pattern such as a chess board is determined to be a backlight situation, and there is a possibility that correction not suitable for the subject to be photographed may be performed.
Japanese Patent No. 4011073 (page 4-5)

  As described above, the conventional backlight determination method has a problem that the backlight condition cannot be appropriately determined depending on the position and condition of the subject, and preferable image data cannot be provided.

The present invention has been made in view of such points, appropriately determine a backlight situation, and an object thereof is to provide the images data processing equipment that can provide desirable image data.

Claim 1 image data processing apparatus described, the dark portion selection reference value, and the reference value setting means for setting a dark region speed reference value, divides the images data into a plurality of regions, the luminance data of the respective areas Area brightness calculation means for calculating and the brightness data is smaller than the dark area selection reference value, and the number of the adjacent areas is calculated as the number of dark area areas, and the number of dark area areas is equal to or greater than the dark area number reference value And determining means for determining the backlight condition, a backlight determining apparatus for determining whether or not the image data was captured in the backlight condition, and correcting the luminance data of each pixel constituting the image data. Brightness correction means for performing correction of the brightness data of each pixel in accordance with the conversion characteristics according to the magnitude of the brightness data of each pixel, the brightness correction means, and the first conversion characteristics and the first conversion For characteristics And a second conversion characteristic in which the correction amount for increasing the luminance data of the dark pixel having a small degree is set to be small, and the backlight determination device does not determine the image data as a backlight condition, When the luminance data of each pixel is corrected according to the first conversion characteristic, and the backlight determination device determines that the image data is in a backlight state, the luminance data of each pixel of the image data is converted to the second conversion characteristic. The correction is made according to the following .

In this configuration, it is appropriately determined whether or not the image data is captured under backlight conditions regardless of the position of the subject. Also, the luminance correction means corrects the luminance data of each pixel, prevents the dark part from being crushed dark, and provides preferable image data. When the backlight determination device determines that the backlight condition is present, inappropriate correction due to an excessively large correction amount is suppressed by suppressing the correction of the luminance data in the dark part. Therefore, it is possible to appropriately process the image data, and it is possible to provide preferable image data.

The claim 2 image data processing apparatus, wherein in the image data processing apparatus according to claim 1, wherein, the reference value setting means sets the dark portion similar reference value, determining means, the luminance data than the dark portion selection reference value The number of the areas which are small and the difference in the luminance data between the adjacent areas is smaller than the dark part similarity reference value is calculated as the number of dark part areas, and the number of dark part areas is equal to or more than the dark part area number reference value. In some cases, a backlight situation is determined.

  In this configuration, it is appropriately determined whether or not the image data is captured in the backlight condition from the dark part situation.

Image data processing apparatus according to claim 3, wherein, in the image data processing apparatus according to claim 2, wherein the reference value setting means sets the light portion selected reference value, the bright portions similar reference values, and a bright region speed reference value The determination means calculates the number of the areas where the brightness data is larger than the bright part selection reference value and the difference of the brightness data from adjacent areas is smaller than the bright part similarity reference value. The number of bright areas is determined, and when the number of bright areas is equal to or larger than the reference number of bright areas and the number of dark areas is equal to or larger than the dark area count reference value, the backlight condition is determined.

  In this configuration, it is appropriately determined whether or not the image data is captured in the backlight condition based on the situation of the dark part and the bright part.

4. The image data processing apparatus described, in the image data processing apparatus according to any one claim 1 to 3, the reference value setting means may be set by calculating the proper luminance data from the image data, the suitable brightness An exclusion condition value larger than the data is set, and the determination means does not determine that the backlight condition is present when there is no luminance data area equal to or greater than the exclusion condition value.

In this configuration, it is appropriately determined whether or not the image data is captured under the backlight condition by excluding the case where there is no bright part with large luminance data .

  According to the present invention, it is possible to appropriately determine whether or not image data is captured under backlight conditions regardless of the position of the subject. Therefore, it is possible to appropriately process image data and provide preferable image data.

Hereinafter, an embodiment of the images data processing equipment of the present invention with reference to the drawings.

  In FIG. 2, reference numeral 10 denotes a camera. The camera 10 is a so-called digital still camera, and records image data photographed by the imaging means 12 on a recording medium 14 as digital data. The camera 10 includes a case 20.In this case 20, an aperture device 21, a lens 22, an image pickup means 12 that is a charge coupled device (CCD), a double correlation sampling circuit (CDS) 24, an amplification circuit 25, Analog-to-digital converter (A / D) 26, first image data memory 27, image data processing unit 28 constituting brightness correction means, second image data memory 29, compression / decompression circuit 31, timing generator 32, memory controller 33, a backlight determination device, a reference value setting unit, a region luminance calculation unit, a determination unit, a control unit 34 having a CPU constituting a luminance correction unit, a power supply circuit 35, and the like are housed. The control means 34 is connected to a working RAM, a ROM that records preset values, and the like. Further, the case 20 includes a display device 36 that is a liquid crystal panel (LCD), an operation unit 37, a recording medium mounting unit 38, a battery storage unit 39 that stores a battery (not shown), an input / output unit 41, and the like. . In the camera 10, the control unit 34 controls the diaphragm device 21, the lens 22, the amplification circuit 25, the image data processing unit 28, the compression / decompression circuit 31, the timing generator 32, and the memory controller 33. Further, the control means 34 controls the imaging means 12, the double correlation sampling device circuit 24, and the analog / digital converter 26 via the timing generator 32. Further, the control means 34 controls the first image data memory 27 and the second image data memory 29 via the memory controller 33.

  The camera 10 converts the light of the subject captured through the aperture device 21 and the lens 22 into an electrical signal by the imaging means 12. The electric signal output from the image pickup means 12 is subjected to signal processing by the double correlation sampling circuit 24, and the luminance of a large number of R (red), G (green), and B (blue) pixels by the analog-digital converter 26. It is converted into image data consisting of data (gradation values). This image data is temporarily recorded in the first image data memory 27. The image data temporarily recorded in the first image data memory 27 is processed by the image data processing unit 28 with respect to luminance, that is, gradation and white balance, and is displayed on the display device 36, and the second image data. Recorded in the memory 29. Further, the image data recorded in the second image data memory 29 is compressed by the compression / decompression circuit 31 and recorded as image data on the recording medium 14.

  In the camera 10, the state is switched by the user operating the operation unit 37. For example, in the preview mode before the main photographing, the imaging unit 12 continuously outputs an electrical signal at a constant interval, The image data output from the data processing unit 28 is displayed on the display device 36. Then, the control means 34 controls the aperture of the aperture device 21 so as to obtain an appropriate exposure as the automatic exposure control (AE), sets the exposure time of the image pickup means 12, and performs automatic focus control (AF). As described above, the lens 22 is controlled so that the elephant of the target subject is focused by the imaging means 12.

  Here, the control means 34 analyzes the image data recorded in the first image data memory 27, and performs automatic exposure control and automatic focus control automatically or according to the operation of the operation unit 37 by the operator. Do. That is, image data composed of luminance data of a large number of pixels is divided into a predetermined number of vertical and horizontal blocks, an average value of RGB luminance data is calculated for each block, and luminance data for each block is calculated. The luminance data of this block is used to determine the shooting situation (scene), the position of the subject, etc., calculate and set appropriate luminance data, and perform appropriate automatic exposure control according to the subject.

  Then, when the user half-presses the shutter of the operation unit 37, the camera is ready for shooting, the exposure and focus are confirmed, and the user fully presses the shutter of the operation unit 37 to The imaging unit 12 performs exposure for a predetermined exposure time controlled by the control unit 34, outputs an electrical signal based on the accumulated charges, and passes through the first image data memory 27 to the image data processing unit 28. The image data processed in the above is recorded on the recording medium 14 via the second image data memory 29.

  Here, the control means 34 analyzes the image data recorded in the first image data memory 27 or the second image data memory 29 to grasp the status of the image data, and automatically or by the operator's operation. In accordance with the operation of the unit 37, the image data processing unit 28 is controlled to perform appropriate image data processing and provide preferable image data.

  The image data processing method performed by the control means 34 is, for example, a non-linear gradation correction in which the input / output characteristics for correcting the luminance data of each pixel according to the magnitude of the luminance are non-linear gradation correction. The first conversion characteristic which is so-called normal brightness-up processing to increase the brightness data of the region where the brightness is low (lightness is low), that is, the dark portion, and the case where the control means 34 analyzes the image data and determines the backlight condition. A second conversion characteristic that is backlight correction applied in the processing is applied. The second conversion characteristic is set so as to reduce the degree to which the brightness data of the dark portion is increased and to output more natural image data as compared with the first conversion characteristic.

  Next, a description will be given of a backlight determination method in which the control means 34 determines whether or not image data has been captured under backlight conditions.

  In this backlight determination method, as in automatic exposure control, image data composed of luminance data of a large number of pixels is divided into a predetermined number of blocks in the vertical and horizontal directions. For example, image data obtained by photographing the image shown in FIG. 4 is divided into blocks (A shown in FIG. 4) as 12 × 8, that is, horizontal 12 and vertical 8 areas, as shown in FIG. An average value is calculated, and luminance data for each block is calculated from the average value of RGB. Note that when calculating the average value of RGB luminance data, calculation is performed including so-called saturated pixels in which the luminance data has the maximum value.

  The luminance data for each block is expressed by appropriate luminance data, that is, a luminance difference (delta [EV]) that is a deviation from the appropriate exposure. In this case, a block with appropriate exposure is set to 0 EV, a block photographed underexposed is represented by a negative numerical value, and a block photographed with overexposure is represented by a positive numerical value. EV is a so-called exposure value, and the value increases or decreases by 1 each time the exposure time is halved or doubled.

  Next, as shown in FIG. 1, white cluster detection, black cluster detection, cluster number determination, and exclusion condition determination are performed.

  First, as black-and-white cluster detection (step 1), when blocks having a small luminance or lightness exist in a predetermined number or more, a black cluster (Cb shown in FIG. 4) is obtained, and the number of blocks included in the black cluster is counted. When blocks having a large luminance, that is, lightness exist in a predetermined number or more, a white cluster (Cw shown in FIG. 4) is used, and the number of blocks included in the white cluster is counted.

  As white cluster detection for detecting bright blocks, white clusters are detected according to the following conditions W1 to W3, and the number of blocks included in these white clusters is counted.

  W1. Brightness data larger than a preset bright part selection reference value, that is, a bright block having a large luminance difference from proper exposure. Here, the luminance difference delta> +1.5 [EV].

  W2. Next, for bright blocks, it is determined whether the surrounding blocks are similarly bright. That is, even if it is a bright block, an isolated block in which luminance data is similar to an adjacent block, that is, a block having a luminance difference smaller than a preset bright part similarity reference value is excluded. Here, if at least one block having a difference in luminance difference delta within ± 0.6 [EV] is adjacent, the number of blocks is counted as a similar cluster that satisfies the condition.

  W3. Further, regarding the blocks excluded as isolated under the condition of W2, if the adjacent blocks are darker than the preset bright part exception reference value, the number of blocks is counted as satisfying the condition. Here, the condition is satisfied if the luminance difference delta with at least one adjacent block is greater than 1.5 [EV] and at least one dark block (luminance difference delta <-1.5 [EV]) is adjacent. As above, count the number of blocks.

  In this way, the number of blocks satisfying the conditions of W1 and W2 and the number of blocks satisfying the conditions of W1 and W3 are totaled and recorded as the number of bright areas.

  As black cluster detection for detecting dark blocks, black clusters are detected according to the following conditions B1 to B3, and the number of blocks included in these black clusters is counted.

  B1. Luminance data smaller than a preset dark area selection reference value, that is, a dark block having a large luminance difference from proper exposure. Here, it is assumed that the luminance difference delta <−3.0 [EV].

  B2. Next, for dark blocks, it is determined whether the surrounding blocks are dark as well. That is, even in the case of a dark block, an isolated block in which luminance data is similar to an adjacent block, that is, a block having a luminance difference smaller than a preset dark part similarity reference value does not exist. Here, if at least one block having a difference in brightness difference delta within ± 0.4 [EV] is adjacent, the number of blocks is counted as a similar cluster that satisfies the condition.

  B3. Further, regarding the blocks excluded as isolated under the condition of B2, if the adjacent blocks are brighter than the preset dark part exception reference value, the number of blocks is counted as satisfying the condition. Here, the condition is satisfied if the luminance difference delta with at least one adjacent block is larger than 3.0 [EV] and at least one bright (luminance difference delta> −3.0 [EV]) block is adjacent. As above, count the number of blocks.

  In this way, the number of blocks satisfying the conditions of B1 and B2 and the number of blocks satisfying the conditions of B1 and B3 are totaled and recorded as the number of dark areas.

  Next, as cluster size determination (step 2), it is determined whether the number of bright areas and the number of dark areas are equal to or greater than a predetermined number, respectively.

  Here, the number of blocks included in one or a plurality of black clusters is a reference value for the number of dark areas set in advance as a ratio to the total number of blocks, for example, ¼ or more of the total number of blocks, that is, 24 or more. In addition, the number of blocks included in one or a plurality of white clusters is a reference value for the number of bright areas set in advance as a ratio to the total number of blocks, for example, 1/16 or more of the total number of blocks, that is, 6 or more in this case. It is assumed that the conditions for determining the cluster size are satisfied.

  Next, when the condition for determining the cluster size is satisfied, it is determined as a determination of the exclusion condition (step 3) whether there is a block of luminance data larger than a predetermined value, that is, a predetermined exclusion condition value.

  Here, the control means 34 calculates and sets an exclusion condition value that is an adjustment value according to the appropriate luminance data calculated from the image data, for example, an exclusion condition value that is 2.0 [EV] larger than the appropriate luminance data. In this case, if there is no block with brightness difference delta> +2.0 [EV] or more from the appropriate exposure by looking at the histogram of the brightness data, it is not judged as a backlight situation and the brightness difference from the appropriate exposure. When there is a block of delta> +2.0 [EV] or more, it is determined that the backlight is in the backlight state.

  When the above conditions are satisfied as described above, it is determined that the backlight is in a backlight state (step 4), and the gradation correction is performed by applying the backlight correction, that is, the second conversion characteristic to the image data (step 5). If the above condition is not satisfied, the backlight condition is not judged (step 6), and normal correction, that is, gradation correction is performed by applying the first conversion characteristic to the image data (step 7).

  As described above, according to the present embodiment, in the image data processing method for determining the backlight condition and correcting the dark portion of the image data by post-processing, when pixels having small luminance data exist in a cluster of a predetermined range or more. By determining the backlight condition, and further determining the backlight condition when pixels with small luminance data are present in a cluster of pixels greater than or equal to the first predetermined range and pixels with large luminance data are equal to or greater than the second predetermined range. From the situation of the dark part and the bright part, regardless of the position of the subject, that is, not only when the subject exists at the center as shown in FIG. 5, but also when the subject is not at the center as shown in FIG. It is possible to appropriately determine whether or not the image data is captured in a backlight condition.

  Also, the backlight condition intended for correction in the image processing can be detected by counting the blocks that have become a solid rather than simply counting the number of blocks.

  Further, by counting blocks adjacent to blocks with a large luminance difference together with the blocks that have become agglomerated, it is possible to detect a backlight condition intended for correction in image processing.

  Further, in the determination of the backlight condition, it is possible to appropriately determine whether or not the image data is captured under the backlight condition by excluding the case where there is no bright part with large luminance data.

  Therefore, depending on whether or not the image data was captured in a backlight situation, the luminance correction processing (brightness processing) of the image data can be appropriately performed, and preferable image data can be provided.

  As for the correction of the image data, it is possible to provide preferable image data by correcting the luminance data of each pixel to prevent the dark part from being crushed darkly. And when it is determined that the backlight is in the backlit condition, by suppressing the correction of the brightness data in the dark area, the inappropriate correction due to the correction amount becoming too large is suppressed, and the specific color and noise are prevented from being strong. And preferable image data can be provided.

  Further, since the luminance data blocking process for automatic exposure control can be used for the determination of the backlight condition, the program and configuration can be simplified and the manufacturing cost can be reduced.

  Each reference value serving as a threshold can be a value recorded in advance in the ROM as a constant value, or can be set linearly or non-linearly according to appropriate luminance data calculated from image data.

  Further, the camera 10 includes a flash (not shown), so that the backlight correction after shooting can be performed together with the backlight correction when shooting with the flash.

  Further, the determination of the backlight condition and the correction of the image data may be automatically performed immediately after shooting, or may be appropriately applied to each image data by the operation of the operator.

  Furthermore, the present invention can be applied to a digital video camera, and the determination of backlight status and the correction of image data can be applied to each image data constituting a moving image.

  The present invention can be applied to, for example, a digital still camera.

It is a flowchart explaining operation | movement of the camera provided with the backlight determination apparatus and image data processing apparatus which show one embodiment of this invention. It is a block diagram of a camera same as the above. It is explanatory drawing of the image image | photographed with a camera same as the above. It is explanatory drawing of operation | movement of a backlight determination apparatus same as the above. It is explanatory drawing of operation | movement of a backlight determination apparatus same as the above.

28 Image data processing section constituting luminance correction means
34 Control means constituting backlight determination device, reference value setting means, area luminance calculation means, determination means, and luminance correction means

Claims (4)

Dark portion selection reference value, and the reference value setting means for setting a dark region speed reference value, the images data is divided into a plurality of regions, said a region luminance calculation means for calculating the luminance data of each area, the dark area selection Judgment that the backlight condition is determined when the brightness data is smaller than the reference value and the number of the adjacent areas is calculated as the number of dark area areas, and the number of dark area areas is equal to or greater than the dark area area reference value. Means for determining whether or not the image data was captured in a backlight condition;
Brightness correction means for correcting the brightness data of each pixel constituting the image data,
The brightness correction means includes
The luminance data of each pixel is corrected according to the conversion characteristics corresponding to the magnitude of the luminance data of each pixel, and the first conversion characteristic and the luminance data of the dark part pixel having a low luminance are increased with respect to the first conversion characteristic. A second conversion characteristic in which a correction amount to be set is set to be small,
When the backlight determination device does not determine the image data as a backlight condition, the luminance data of each pixel of the image data is corrected according to the first conversion characteristic,
When the backlight determination device determines that the image data is in a backlight state, the luminance data of each pixel of the image data is corrected according to the second conversion characteristic.
An image data processing apparatus . The reference value setting means sets the dark part similarity reference value,
The determination means calculates the number of the areas where the brightness data is smaller than the dark part selection reference value and the difference of the brightness data from the adjacent areas is smaller than the dark part similarity reference value to obtain the number of dark part areas. The image data processing apparatus according to claim 1, wherein when the number of dark area areas is equal to or greater than the dark area area reference value, it is determined that the backlight condition is present. The reference value setting means sets a bright part selection reference value, a bright part similarity reference value, and a bright part region number reference value,
The determination means calculates the number of the areas where the brightness data is larger than the bright part selection reference value and the difference of the brightness data between adjacent areas is smaller than the bright part similarity reference value. The number of areas is determined, and when the number of bright areas is equal to or greater than the reference value for the number of bright areas and the number of dark areas is equal to or greater than the reference value for the number of dark areas, the backlight condition is determined. Image data processing apparatus . The reference value setting means calculates and sets appropriate luminance data from the image data, sets an exclusion condition value larger than the appropriate luminance data,
4. The image data processing apparatus according to claim 1, wherein the determination unit does not determine a backlight condition when there is no luminance data area equal to or greater than the exclusion condition value . 5.

Images