JP4063592B2 - Digital camera - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a digital camera and an image processing method thereof.
[0002]
[Prior art]
FIG. 8 is an internal block diagram of a conventional digital camera (see Japanese Patent Application Laid-Open No. 2001-313680). The light reflected by the subject is focused on the light receiving surface of the CCD (33), which is an image sensor, by the objective lens (30). The CCD (33) is controlled in exposure and output timing by the drive circuit (35), converts incident light into an analog electric signal and outputs it. That is, when the luminance level of incident light is high, the level of the electrical signal is also increased. The output signal from the CCD (33) is input to an analog signal processing circuit (34), subjected to sampling hold and gain adjustment, and then converted to a digital signal by an A / D converter (38), and then to a control circuit ( It is stored in the image memory (4) or the memory card (9) via 2). The image in the image memory (4) or the memory card (9) is converted into an analog image by the D / A converter (80) via the control circuit (2) and then displayed on the liquid crystal panel (8).
[0003]
When still images of a subject are continuously shot, a blurred image may be taken due to the camera shake of the photographer or the movement of the subject. In this way, since the image with low resolution is poor in appearance, the photographer checks the images continuously taken from the memory card (9) one by one after photographing the subject and finds the blurred image. Then, the operation button (not shown) is operated to erase it via the control circuit (2). Further, in the case of a moving image in which the movement of the subject is photographed, the moving image portion is erased after all the moving images are once stored in the memory card (9).
[0004]
[Problems to be solved by the invention]
In a conventional digital camera, when a blurred image is taken, it is necessary to manually delete the image each time.
In view of this point, Japanese Patent Laid-Open No. 10-174027 discloses a configuration indicated by a one-dot chain line in FIG. This is because the control circuit (2) is provided with a camera shake detection sensor (16) via an A / D converter (15), and when the camera shake detection sensor (16) detects a camera shake, the control circuit (2) is notified accordingly. The control circuit (2) records an image indicating blurring on the blurred image stored in the image memory (4) or the memory card (9). However, while this makes it convenient to search for blurred images, it still has to be erased manually. In view of the fact that the subject image does not focus on the light receiving surface of the CCD (33) due to camera shake or the like, the luminance level decreases and the level of the high frequency component, which is a high frequency component in the video signal, decreases. The idea was to automatically detect a blurred image from the level of the high frequency component.
An object of the present invention is to provide a digital camera that automatically erases a blurred image.
[0005]
[Means for solving the problems]
The image memory (4) is connected to a face feature point extraction circuit (70) that extracts a feature portion of a human face as a subject, and is divided into a plurality of blocks, and a digital image is stored in each block.
The control circuit (2) searches the high frequency component of the facial feature part included in the luminance signal of the subject image extracted by the facial feature point extraction circuit (70), and the digital image having the high frequency component above a certain level Connected to the discrimination circuit (5) to search for
The control circuit (2) erases the digital image whose high frequency component is determined to be less than a certain level by the discrimination circuit (5), and stores the image having the high frequency component equal to or higher than the certain level in the image memory (4).
[0006]
[Action and effect]
The discrimination circuit (5) searches the high frequency component of the facial feature part included in the luminance signal of the subject image extracted by the facial feature point extraction circuit (70) , and obtains a digital image having a high frequency component above a certain level. look for. Therefore, even if an image with a blurred face is taken, if the discrimination circuit (5) determines that the high frequency component of the image is below a certain level, the control circuit (2) erases the image. Only digital images whose high frequency components are above a certain level continue to be stored in the image memory (4). That is, since the blurred image is automatically erased, it is not necessary to manually erase the blurred image, and the usability is improved.
[0007]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, an example of the present invention will be described in detail with reference to the drawings.
1A is a perspective view of the camera body 1, and FIG. 1B is a perspective view of FIG. 1A viewed from behind. On the camera body (1), in addition to the shutter button (10) and strobe (70), there is provided a changeover switch (SW) which is operated when setting the automatic saving and erasing mode described later. The camera body (1) is provided with an opening (11) into which a memory card (9) capable of storing a digital image is inserted. The taken image can be confirmed on a liquid crystal panel (8) provided on the back of the camera body (1). This example is characterized in that the blurred image is automatically erased.
[0008]
FIG. 2 is an internal block diagram of the digital camera of this example. A digital image signal is output from the autofocus mechanism (3) including the objective lens (30). The digital image signal is separated and converted into a luminance signal and a color difference signal by the digital signal processing circuit (6), and color correction is performed. Is input to the control circuit (2).
The control circuit (2) includes a ROM (20) in which programs are stored, a RAM (21) in which characters and characters to be displayed on the liquid crystal panel (8) are stored, and an image memory (4) in which digital images are stored. It is connected. The control circuit (2) is connected to the liquid crystal panel (8) via a D / A converter (80) and a memory card (9) via a compression / decompression circuit (90). A discriminating circuit (5) connected to the control circuit (2) is connected to the image memory (4), and the discriminating circuit (5) stores a predetermined level value of the high frequency component in advance. ) Is set to the automatic save and erase mode, the discrimination circuit (5) determines whether or not the integral value of the high frequency component contained in the luminance signal of the image in the image memory (4) is above a certain level. Determine.
[0009]
FIG. 3 is a block diagram showing the internal configuration of the autofocus mechanism (3) (see Japanese Patent Laid-Open No. 4-84572). The objective lens (30) facing the subject is provided with a rack (31), and the rack (31) is provided to be close to and away from the subject by a motor (M) via a gear train (32). . Electric power is supplied to the motor (M) from the drive circuit (35). The light from the objective lens (30) is converted into an electrical signal by the CCD (33), subjected to sampling hold and gain adjustment by the analog signal processing circuit (34), and then a high pass filter (HPF) (36). The high-frequency component, which is a high-frequency component, is extracted at, integrated by the integration circuit (37), and then converted into a digital signal by the A / D converter (38). The digitized integrated value is input to the control circuit (2), one terminal of the comparator (69), and the memory (39). The image signal from the analog signal processing circuit (34) is digitally converted by the A / D converter (68) and is also supplied to the digital signal processing circuit (6).
[0010]
The other terminal of the comparator (69) is inputted with a level that is an integral value of the high frequency component taken at the previous imaging by the memory (39). The comparator (69) outputs a high signal to the drive circuit (35) when the current level is higher, and outputs a low signal to the drive circuit (35) when the current level is lower. The drive circuit (35) supplies a shutter pulse to the CCD (33) and energizes the motor (M) according to the output signal of the comparator (69) to move the objective lens (30) along the optical axis L. .
FIG. 4 is a graph showing the relationship between the level, which is the integral value of the high frequency component, and the position of the objective lens (30) on the optical axis. The initial position of the objective lens (30) is defined as position 1 in FIG. The level at this time is stored in the memory (39). The drive circuit (35) moves the objective lens (30) to the position 2 along the optical axis L. Since the level of the high frequency component at this time is larger than the level at the position of 1, the comparator (69) outputs a high signal to the drive circuit (35), and the memory (39) is at the position of 2. The level at is stored. The drive circuit (35) further moves the objective lens (30) in the same direction, and the comparator (69) continues to compare the digital values.
[0011]
When the objective lens 30 reaches the position 4 in FIG. 4, the comparator 69 outputs a low signal because the level of the high-frequency component at the previous position 3 is greater. The drive circuit (35) returns the objective lens (30) to the position 3 and stops energization of the motor (M) at the position 3. Thereby, the objective lens (30) stops at the position where the level of the high frequency component is the largest, and the image of the subject is focused on the CCD (33).
If the initial position of the objective lens 30 is the position 4 in FIG. 4, the comparator 69 outputs a low signal when the objective lens 30 is moved toward the position 5. The circuit (35) reverses the motor (M) to move the objective lens (30) to position 3. The method of focusing the subject image on the CCD (33) in this way is also called a mountain climbing method. The CCD (33) has an auto iris function, and the level of the high frequency component is searched within a certain level limited by the auto iris function.
[0012]
As shown in FIG. 5, the image memory (4) is divided into a plurality of blocks (40) and (40). Each block (40) includes one image, that is, one frame image and the image. The integrated value of the high-frequency component is stored.
The control circuit (2) is connected to each block (40) of the image memory (4) and erases the digital image for each block. In FIG. 5, the number of the blocks (40) is eight, but this is for convenience of explanation and is not limited to this number.
The image memory (4) is formed of, for example, a DRAM. As is well known, the DRAM has a memory cell composed of a capacitor and a MOS transistor. When the digital image is continuously stored in the image memory (4), the refresh voltage is continuously supplied to hold the charge. When erasing a digital image, if the power supply of the refresh voltage is stopped, the capacitor is discharged and the digital image is erased.
[0013]
(Auto save and delete mode)
When the user operates the changeover switch (SW), the program is read from the ROM (20), and the camera body (1) is set to the automatic save and delete mode. In this mode, the blurred image is automatically erased. Note that a menu may be displayed on the screen of the liquid crystal panel (8), and the automatic save and delete modes may be set from the menu. Hereinafter, the operation procedure in the automatic saving and erasing mode will be described with reference to the flowchart of FIG. When the photographer presses the shutter button (10) with the changeover switch (SW) turned on (S1), the subject image is converted into a digital image by the digital signal processing circuit (6), and the control circuit ( 2). The control circuit (2) stores the digital image in each block (40) of the image memory (4) together with the integral value of the high frequency component supplied from the autofocus mechanism (3) (S2).
[0014]
The discrimination circuit (5) judges whether or not the value of the high frequency component stored together with the digital image in each block (40) is equal to or higher than a certain level (S3). If it is above a certain level, this is notified to the control circuit (2), and the control circuit (2) continues to supply power to the block (40) and stores the digital image in the image memory (4) as it is (S4). ). The image stored in the image memory (4) is displayed on the liquid crystal panel (8) or compressed by the compression / decompression circuit (90) and stored in the memory card (9).
If the value of the high frequency component is less than a certain level, the discrimination circuit (5) informs the control circuit (2) that the control circuit (2) cuts off the energization to the block (40) and the digital circuit. The image is erased (S5).
[0015]
In the image memory (4), only digital images whose high frequency component values are above a certain level are stored. That is, since the image quality is low and the image with a low luminance level is automatically erased, it is not necessary to manually erase the blurred image, which improves usability. When there are a plurality of digital images having a high-frequency component value of a certain level or higher, only the digital image having the highest high-frequency component level may be left. That is, even if a plurality of still images of the same subject are left, the photographer may want only one with the highest resolution. In this case, only the digital image with the highest level of the high frequency component is left.
In particular, the function of leaving only the digital image with the highest level of the high frequency component is effective in the continuous shooting mode in which digital images of the same subject are automatically and continuously photographed. That is, when different subject images are taken, if only the digital image having the highest level of the high frequency component is left, there is a risk that the image desired by the photographer will be erased. On the other hand, in the continuous shooting mode, since an image of a subject with the least blur is usually desired, if only a digital image having the highest level of the high frequency component is left, the utility is high.
[0016]
In the above example, the high frequency component (36) is extracted from the luminance signal of the subject image, and the discrimination circuit (5) determines whether the integrated value of the high frequency component is equal to or higher than a certain level. . However, the compression / decompression circuit 90 may compress still images by an image compression method called JPEG (joint photographic experts group). JPEG often uses lossy compression that truncates elements that are not visually important, and uses an encoding method called discrete cosine transform (DCT).
This DCT expresses a planar luminance change distribution by frequency components constituting the DCT, and the discrimination circuit (5) determines whether or not the high frequency component of this frequency is above a certain level. Good. In other words, the value of the high frequency component may be obtained not from the autofocus mechanism (3) but from the compression / expansion circuit (90).
Further, if there are many high frequency components in the image, the file size stored in the image memory (4) becomes large. Therefore, an image to be deleted and an image to be left may be selected from this file size.
[0017]
The autofocus mechanism (3) generally focuses near the center of the objective lens (30). Accordingly, in the automatic saving and erasing mode, the discrimination circuit (5) determines whether or not the center portion of the image is blurred. It should be noted that the image recognition area may be limited automatically or manually, and the center portion may be focused on image recognition. Further, the autofocus mechanism (3) focuses the subject image on the CCD (33) from the level of the high frequency component included in the luminance signal, but instead, infrared rays are emitted from the camera body (1). A method of irradiating the subject and measuring the distance may be used. In this case, as described above, the value of the high frequency component is obtained from the compression / decompression circuit (90).
Even if the autofocus mechanism (3) is not provided, the center of the image may be focused by manual operation. The contents of the present application can also be used for a pan focus machine in which the aperture of the objective lens (30) is reduced to increase the depth of field.
[0018]
(Face feature point detection)
When the subject is a human face, it is preferable to reduce the shake of the face. In this case, it is conceivable to reduce the blur by specifying the face area from the skin color of the face and detecting the blurring of the eyes and nose which are characteristic parts in the face area. Japanese Patent Laid-Open No. 11-281661 discloses a technique for identifying a face area from the skin color of the face and detecting eyes and noses in the face area. Further, whether or not to select a mode in which the eye and nose shake in the face area are detected with priority can be selected from the menu screen of the liquid crystal panel (8).
The one-dot chain line in FIG. 2 is obtained by adding the configuration disclosed by the applicant in Japanese Patent Laid-Open No. 1-281661 to this example, and the operation is shown in the flowchart of FIG. A face feature point extraction circuit (70) is connected to the image memory (4) via a skin color extraction circuit (7), and the face feature point extraction circuit (70) is connected to a discrimination circuit (5).
The skin color extraction circuit (7) histograms the image read from the image memory (4) using the HSV color system defined as a color system that is relatively close to human color vision, and calculates the peak value in the skin color region. The reference skin color hue value is used. A difference between the reference flesh color hue value and the pixel of the image memory (4) is obtained, a histogram is created, and a flesh color region is determined by binarization at a predetermined threshold level. This HSV color system is composed of three elements of H (hue), S (saturation), and V (lightness), and H and V are independent.
The skin color extraction circuit (7) reads a digital image from the image memory (4) (S10), and sets the determined skin color area as a face area (S11). If the skin color area is not determined, the image is read again from the image memory (4).
[0019]
The face area data is sent to a face feature point extraction circuit (70). In the face feature point extraction circuit (70), approximate shapes of eyes and nostrils are stored in advance as templates. The face feature point extraction circuit (70) searches whether there is an element having a shape similar to the template in the face area determined by the skin color extraction circuit (7) (S12). If an element having a shape similar to the template is found, the position of the element is transmitted to the discrimination circuit (5) (S13), and if not found, the image is read from the image memory (4) again. The discriminating circuit (5) intensively observes the high-frequency component at the position of the element, and judges whether or not it is above a certain level. This can reduce facial blur.
[0020]
The above description of the embodiments is for explaining the present invention, and should not be construed as limiting the invention described in the claims or reducing the scope thereof. In addition, the configuration of each part of the present invention is not limited to the above embodiment, and various modifications can be made within the technical scope described in the claims.
[Brief description of the drawings]
FIG. 1A is a perspective view of a camera body, and FIG. 1B is a perspective view of FIG.
FIG. 2 is an internal block diagram of the digital camera of this example.
FIG. 3 is a block diagram showing an internal configuration of an autofocus mechanism.
FIG. 4 is a graph showing a relationship between a level, which is an integral value of a high frequency component, and a position on the optical axis of an objective lens.
FIG. 5 is a diagram showing blocks in an image memory.
FIG. 6 is a flowchart showing an operation procedure in an automatic saving and erasing mode.
FIG. 7 is a flowchart showing a procedure for detecting feature points of a face.
FIG. 8 is an internal block diagram of a conventional digital camera.
[Explanation of symbols]
(2) Control circuit
(4) Image memory
(5) Discrimination circuit
(30) Objective lens
(70) Facial feature point extraction circuit

Claims (3)

An objective lens (30) for condensing image light from the subject;
Based on the brightness level of the subject image focused by the objective lens (30), an image sensor that outputs an electrical signal, an image memory (4) that can store a plurality of digitally converted image signals, and an image memory In a digital camera comprising a control circuit (2) connected to (4),
The image memory (4) is connected to a face feature point extraction circuit (70) for extracting a feature portion of a human face as a subject, and is divided into a plurality of blocks, and a digital image is stored in each block.
The control circuit (2) searches the high frequency component of the facial feature included in the luminance signal of the subject image extracted by the facial feature point extraction circuit (70), and has a high frequency component of a certain level or higher. Connected to the discrimination circuit (5) to search for
The control circuit (2) deletes the digital image whose high frequency component is determined to be less than a certain level by the discrimination circuit (5), and stores the image having the high frequency component above the certain level in the image memory (4). A digital camera characterized by   The digital camera according to claim 1, wherein the control circuit (2) stores only an image having a highest level high frequency component among images having a high frequency component of a certain level or higher.   The digital camera according to claim 2, wherein the mode for photographing the subject is continuous shooting.

Images