KR100619306B1 - Abstraction method for bad pixel of digital imaging apparatus - Google Patents

Abstract

The present invention relates to a method for extracting defective pixels of a digital imaging apparatus, in particular, capable of self-correcting defective pixels of an image captured from a camera in the imaging apparatus.

In accordance with an aspect of the present invention, there is provided a method for extracting a defective pixel of a digital imaging apparatus, the method including: operating basic parameters preset for initial setting for a defective pixel test of a camera; When the initial setting is completed according to the basic parameter, capturing a plurality of images by differently adjusting shutter speed values; Extracting luminance components from the plurality of images to obtain a plurality of sample images; And determining a dead pixel when a value of a luminance component of a specific coordinate in all of the plurality of sample images coincides with a white value.

Camera, Bad Pixel, Dead Pixel, Hot Pixel

Description

Abstraction method for bad pixel of digital imaging apparatus

1 is a block diagram showing an apparatus for extracting defective pixels of a digital camera device according to an embodiment of the present invention.

2 is a flowchart illustrating a bad pixel test method of a digital camera device according to an exemplary embodiment of the present invention.

3 is a flowchart illustrating a method for extracting a bad pixel of a digital camera device according to an exemplary embodiment of the present invention.

4 shows an example of pixel mapping of a dead pixel according to the present invention;

<Description of the symbols for the main parts of the drawings>

20 ... camera module

30 ... core processor

Camera device driver

32.Camera Buffer

33.Camera register

The present invention relates to a method for extracting defective pixels of a digital imaging apparatus, in particular, capable of self-correcting defective pixels of an image captured from a camera in the imaging apparatus.

In general, imaging devices such as digital cameras, scanners, and camera phones can have image sensors produced as a charge coupled device (CCD), a charge injection device (CID), or a complementary metal-oxide semiconductor (CMOS) as a component. have.

The image sensor consists of "sense" arrays or pixel locations that capture energy from an illuminant and often converts the energy into a specific magnitude, such as an intensity value.

In general, there are two types of bad pixels. The first is shown on an image sensor (CCD, CMOS, etc.) where bad pixels remain in the captured image. The second is on the camera's LCD.

In most cases, the image sensor has any number of defective pixels due to manufacturing or manufacturing errors, which are the dead pixels and hot pixels in the image sensor.

Here, the dead pixel always appears because the pixel of the image sensor is dead, that is, it does not work. This dead pixel is always white because no light is input and cannot be recovered again. Only pixel mapping at the after-sales center makes the bad pixels invisible.

In the pixel mapping, images stored in an image device are downloaded to a computer from an after-sales center and then tested for each image through a bad pixel inspection program. Then, the pixel mapping is performed to extract dead pixels and perform pixel mapping.

In addition, a hot pixel is a phenomenon in which an image sensor receives heat due to long time or short time shooting with the camera turned on for a long time.

It is very difficult to ensure that no pixels in the sensor will be defective during the fabrication / manufacturing of such an image sensor. In addition, the imaging device does not have the ability to check and diagnose defective pixels on its own.

That is, in the conventional imaging apparatus such as a digital camera, a method of downloading a JPEG image of an RGB color to a PC and confirming it with the naked eye or an application program has been used.

In addition, in conventional CIF (352 * 288) or VGA (640 * 480) cameras, the resolution is relatively low, so no consideration has been given to defective pixels. In a "mega-pixel" image sensor, many pixels may be defective. Accordingly, there is a demand for a method in which a user can easily check a defective pixel in a megapixel imaging device having at least one million effective pixels.

In addition, recently, as high quality imaging devices have been spread, users' knowledge and level of cameras have been greatly improved. Even if a service of various functions is added to such an imaging device, if a defective pixel appears in an image captured by a camera, a problem that does not satisfy the needs of the user may occur unless it is solved by itself.

It is a first object of the present invention to provide a method for extracting defective pixels of a digital imaging apparatus, by which a defective pixel can be checked by itself in an imaging apparatus, so that a user can easily test whether the defective pixel is present.

A second object of the present invention is to provide a method for extracting defective pixels of a digital imaging apparatus, which enables self-checking of defective pixels in the imaging apparatus and self-correcting the corresponding pixels.

A third object of the present invention is to provide a method for extracting a defective pixel of a digital imaging apparatus, which can provide a user with coordinates of a dead pixel and a hot pixel found in the imaging apparatus.

Defective pixel extraction method of the digital imaging device according to the present invention for achieving the above object,

Operating a preset basic parameter for initial setting for a bad pixel test of the camera;

When the initial setting is completed according to the basic parameter, capturing a plurality of images by differently adjusting shutter speed values;

Extracting luminance components from the plurality of images to obtain a plurality of sample images;

And determining a dead pixel when a value of a luminance component of a specific coordinate in all of the plurality of sample images coincides with a white value.

Preferably, the method may further include pixel mapping the pixel values of the specific coordinates by using a predetermined number of neighboring pixel values.

The method may further include determining the pixel as a hot pixel when a value of a luminance component of a specific coordinate in all of the plurality of sample images exceeds a threshold.

Preferably, characterized in that the user to confirm the specific coordinates determined as the dead pixel or hot pixel.

Preferably, the operation of the basic parameter for the initial setting is characterized in that it comprises lens light blocking, noise filter off, flash off, resolution maximum setting, camera sensitivity minimum setting.

In addition, the defective pixel extraction method of the digital imaging device according to another embodiment of the present invention,

In the digital imaging apparatus, when a user performs a bad pixel test, initializing camera setting values using basic parameters, and capturing a predetermined number of sample images by capturing different shutter speed values; Determining whether or not the defective pixel is for the luminance component of the same coordinate in the predetermined sample image; As a result of the determination, the corresponding coordinate determined as the defective pixel is shown to the user.

Preferably, the bad pixel determination is a dead pixel if the value of the luminance component at the same coordinates of a predetermined sample image is a white value, and self-correcting the pixel value of the coordinate where the dead pixel is found by the average of neighboring pixel values, The hot pixel may be determined when a value of a luminance component is equal to or greater than a gray value at the same coordinates of the predetermined sample images.

Referring to the accompanying drawings, a method for extracting a defective pixel of an imaging apparatus according to an exemplary embodiment of the present invention as described above is as follows.

Referring to FIG. 1, the camera module 20 is connected to the core processor 30, the camera clock is normally input to the camera module 20 by the camera device driver 31, and the camera is connected by the camera register 33. Normal control is performed to the module 20.

At this time, the output of the camera module 20 is output as a YUV stream composed of luminance (Y) components and chroma (C _B , C _R ) components, and stored in the camera buffer 32 through the camera device driver 31.

At this time, the core processor 30 has the YUV data stored in the camera buffer 32 and checks the defective pixel of the camera and corrects itself.

Here, the camera output is divided into a luminance (Y) component having brightness information and a saturation (C _B , C _R : chrominance) component having color information. One pixel consists of one luminance component and two chroma components. Then, a constant byte (for example, 1 byte) is used for the luminance component per pixel. For example, there is a 4: 2: 0 form having one C _B byte and a C _R byte per 2 * 2 pixel array, and a 4: 2: 2 form having two C _B bytes and a C _R byte. This is a common form of camera output data mounted to an imaging device.

In this case, the luminance component has a value ranging from black (black) 16 to white (235), and the C _B and C _R components representing chroma components have values from 16 to 240. . The luminance component value is used to determine whether a dead pixel is present.

2 is a flowchart illustrating a camera bad pixel extraction method according to an embodiment of the present invention.

Referring to FIG. 2, first, when a camera bad pixel test is started (S31), the camera operates with basic parameters preset for an initial setting of the bad pixel test.

The basic parameters of the defective pixel test are set to minimum lens light blocking (S32), noise filter off (S33), flash off (S34), maximum resolution setting (S35), and ISO (International Standards Organization) corresponding to camera sensitivity. (S36).

Here, the lens light blocking operation (S32) may perform the light blocking by the user by showing a message to the user in a scenario. If the camera supports the automatic lens cap (CAP), the lens light blocking function may be software. Can be done as

Then, the noise filter off and flash off operations S33 and S34 are performed to check the correct defective pixels from the functional problem of the pure camera.

The maximum resolution setting operation S35 is the highest resolution of the current camera. For example, in the case of a 1.3 mega pixel camera of SXGA (1280 * 960), the resolution is set to SXGA resolution.

Thereafter, the shutter speed (or shutter exposure time) is adjusted in various ways (S37) to capture an image, and the shutter speed is multiplied by a predetermined multiple of the basic exposure time to capture several types of images with several exposure times. do. Here, the constant multiple is an integer, for example, by adjusting the basic exposure time to 1 second (S38), 2 seconds (S39) twice the basic exposure time, 4 seconds (S40) four times the basic exposure time, respectively, the image Will be captured.

As such, when each sample image is obtained for each exposure time, all the sample images are all black because the light is completely blocked by the lens.

In addition, sample images captured at different shutter speeds do not perform JPEG encoding as in a conventional digital camera, and internally, a camera buffer, which is the camera output data, is brought and checked in units of a maximum resolution. That is, the maximum resolution size unit is one image sample unit.

For example, in the case of 1.3 mega pixels, when capturing three sample images with different shutter speeds (or exposure times), the data in the camera buffer is read and processed by 1280 * 960 bytes. In this case, the amount of computation can be significantly reduced than in the conventional digital camera method, YUV data in the camera buffer is encoded as a JPEG image and RGB conversion is performed to determine bad pixels.

Thereafter, only the Y component is extracted from each sample image (S41), and when each sample image (# 1, # 2, # 3) is obtained (S42, S44, S45), the bad pixel test routine as shown in FIG. 3 is performed. It is made (S45). Here, each sample image is obtained with three shutter speeds (or exposure times), but three or more or three or less may be used according to the characteristics of the camera. It is also desirable to gap the at least two sample images to detect defective pixels at the same coordinates.

In this case, one luminance component and two chroma components are stored in the camera buffer, but one luminance component for each pixel is considered to check for defective pixels, and the chroma component is excluded. Can be.

Referring to Figure 3 describes the defective pixel test method of the present invention as follows.

First, in FIG. 2, a bad pixel verification routine is performed on each sample image of the luminance component (S51). At this time, the value Y (i, j) of the luminance component at the same coordinates (i, j) in all the sample images # 1, # 2, # 3 is the white value (Y (i, J) = 235). It is determined whether the same as (S52).

In this case, when the luminance component of a specific coordinate is the same as the white value 235 in all the sample images, the corresponding coordinate is determined as a dead pixel (S53), and pixel mapping is performed for self-correction (S54). Here, if it is determined that the dead pixel, the coordinate of the camera where the dead pixel is found may be shown to the user.

That is, when the coordinates of the luminance components of all the sample images are white values 235, the camera sensor transistors of the corresponding coordinates are dead pixels that do not operate. Coordinates determined as dead pixels are self-revisioned through pixel mapping. These self-revisions may be performed automatically or by a user command, and may be correlated to a specific camera vendor. You can do pixel mapping for all dead pixels without

Self-correction is then performed using a median filter using adjacent pixel values including the generated dead pixel.

Referring to this self-correction method, as shown in FIG. 4, when one pixel 42 in the middle of the sample image 40 is a dead pixel, the neighboring pixel 41 value surrounding the dead pixel (Neighborhood Pixel) Values are selected based on the median value, and the median value is mapped to the dead pixel.

For example, in the case of FIG. 4, neighboring pixel 41 values of the dead pixel are 124, 126, 127, 120, 125, 115, 119, and 123, respectively, and the sum of these pixels is the number of neighboring pixels (= 8). Dividing is 122. Then, a value of 122 may be mapped to the coordinate 42 'where the dead pixel is generated. As another example, the sum of the six pixels 124, 127, 120, 125, 115, 123 or 124, 126, 127, 115, 119, 123 may be averaged without deriving the sum of the eight adjacent pixels.

Here, since the dead pixel always has a YUV white 235 value regardless of the shutter speed or exposure time of each sample image, the dead pixel has a great influence on the image except for shooting a very bright background. This pixel mapping technique for dead pixels allows a user to conveniently obtain a self-correcting normal image on the imaging device.

On the other hand, it is checked whether an arbitrary coordinate of each sample image exceeds a preset threshold for hot pixels (S55), and when the threshold is exceeded, the corresponding coordinate is determined to be a hot pixel (S56). (S57). In this case, since the hot pixel is generated by the image sensor receiving heat, the hot pixel is displayed at the same coordinate of all the sample images, but at a coordinate not overlapping with the dead pixel. The threshold is determined halfway between black and white.

In other words, a hot pixel is a pixel that occurs when the camera is used for a long time, and the pixel appears at a predetermined position according to the exposure time. Transistors in CCDs require three colors to represent colors: R, G, and B. When one or more of the RGB transistors fail to represent normal colors, they are called hot pixels.

Here, regardless of the camera CCD manufacturer, when the bad pixel test simulation is taken with the lens cap closed as above, it is determined as a hot pixel if the gray level is above the threshold (110) at a certain place regardless of the exposure time. . In other words, the color value is between black 16 and white 235, and the hot pixel may be about 50% of black and white.

In other words, if a value greater than or equal to gray 110 is determined, it is determined as a hot pixel. For example, if the black color should be 50% of the luminance component regardless of the exposure time, it can be judged as a hot pixel.

 Here, the dead pixel determination step or the hot pixel determination step may be changed regardless of the order in which the steps are performed.

The defective pixel test process as described above may be processed in a one-touch manner through a user interface in the imaging device.

In addition, you can perform not only the bad pixel test process but also the bad pixel test process with one touch, and automatically or deadly check the dead pixel self-calibration by checking the dead pixel and hot pixel coordinates. You can also make a choice. If the automatic correction of the dead pixels is selected, the self-correction is automatically performed through pixel mapping on the captured image, so that the user can always see only the self-corrected image.

So far, the present invention has been described with reference to the preferred embodiments, and those skilled in the art to which the present invention pertains to the detailed description of the present invention and other forms of embodiments within the essential technical scope of the present invention. It can be implemented. Here, the essential technical scope of the present invention is shown in the claims, and all differences within the equivalent range will be construed as being included in the present invention.

As described above, the method for extracting defective pixels of the digital imaging apparatus according to the present invention enables the user to easily check the defective pixels in the apparatus and to self-correct the checked defective pixels in response to the high quality imaging apparatus. There is an effect to be able to deal with defective pixels.

In addition, the defective pixel is checked with the data of the YUV buffer in which the output data of the camera is stored in the imaging apparatus, and the processing amount is considerably reduced by processing only the component of Y, which is a luminance component.

In addition, when a dead pixel is detected in the imaging apparatus, self-calibration may be performed using an intermediate filter, so that a user may easily check and correct defective pixels quickly.

Claims (8)

Operating a preset basic parameter for initial setting for a bad pixel test of the camera; When the initial setting is completed according to the basic parameter, capturing a plurality of images by adjusting shutter speed values differently; Extracting luminance components from the plurality of images to obtain a plurality of sample images; And determining a dead pixel if a value of a luminance component of a specific coordinate in all of the plurality of sample images coincides with a white value. The method of claim 1, And mapping the pixel value of the specific coordinate to a pixel using a predetermined number of neighboring pixel values, if the dead pixel is determined as the dead pixel. The method of claim 1, And determining a corresponding pixel as a hot pixel when a value of a luminance component of a specific coordinate in all of the plurality of sample images exceeds a preset threshold. The method of claim 2 or 3, Defective pixel extraction method of the digital imaging device, characterized in that for confirming the specific coordinates determined as the dead pixel or hot pixel to the user. The method of claim 1, The operation of the basic parameter for the initial setting includes lens light blocking, noise filter off, flash off, resolution maximum setting, camera sensitivity minimum setting. In the digital imaging device, initializing camera setting values using basic parameters in a bad pixel test, and capturing a predetermined number of sample images by gaps with different shutter speed values; Determining whether or not the defective pixel is for the luminance component of the same coordinate in the predetermined sample image; And determining, as a result, the corresponding coordinate determined as the defective pixel to the user. The method of claim 6, The bad pixel determination is a dead pixel when a value of a luminance component is a white value at the same coordinates of a plurality of sample images, and self-corrects a pixel value of a coordinate where the dead pixel is found by an average of neighboring pixel values. A method for extracting a defective pixel of a digital imaging apparatus, characterized in that it is determined as a hot pixel when a value of a luminance component is equal to or greater than a gray value at the same coordinate of a sample image. The method of claim 6, The defective pixel test is performed by the user with a single touch.

Images