JP2004112025A - Defective pixel correction device for solid-state imaging device - Google Patents

Abstract

An object of the present invention is to provide a defective pixel correction device for a solid-state imaging device capable of obtaining a high-quality image.
A defect that outputs a signal of a specific level among output signals of a solid-state imaging device (2) of an all-pixel readout method that has a checkerboard color filter of primary colors and can output all pixels independently and sequentially. It is characterized by comprising a defective pixel detecting means 7 for detecting a pixel and a defective pixel correcting means 8 for correcting a pixel determined as a defective pixel by the defective pixel detecting means 7.
[Selection diagram] Fig. 1

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to a defective pixel correction device for a fixed imaging device that detects and corrects a pixel defect existing in a solid-state imaging device in an imaging device using an all-pixel readout solid-state imaging device.
[0002]
[Prior art]
In general, in a solid-state imaging device formed of a semiconductor, it is known that image quality is deteriorated due to a local crystal defect of the semiconductor. An image defect in which a constant bias voltage is always added to the imaging output according to the amount of incident light appears as a high-luminance white dot on the monitor screen if this image defect signal is processed as it is. Those with low photoelectric sensitivity appear as black dots.
[0003]
Conventionally, detection and correction of a defective pixel as described above is disclosed in, for example, a solid-state imaging device of Patent Document 1. Hereinafter, a conventional defective pixel correction device will be described.
[0004]
FIG. 10 shows the configuration of a conventional defective pixel correction device. This defective pixel correction device includes a lens 1, a field read-out driving solid-state imaging device 20, a preprocessing unit 3, an A / D converter 4, a 1H delay circuit 6, an LPF 12, a defective pixel detection unit 7, a defective pixel correction unit 8, It comprises a luminance signal processing circuit 9 and a color signal processing circuit 10. The defective pixel detecting means 7 extracts a pixel to be determined and its neighboring pixels, calculates an amount of protrusion by using an adder / subtractor and a comparator with respect to these pixel data, and determines a pixel satisfying a condition determined by a predetermined threshold value. Is determined to be a defective pixel. The defective pixel correcting means 8 corrects the defective pixel based on the determination result of the defective pixel detecting means 7.
[0005]
[Patent Document 1]
JP-A-5-41868
[Problems to be solved by the invention]
However, in recent image pickup apparatuses using a solid-state image pickup device such as a CMOS, it is required to obtain a higher-resolution and higher-quality image by using a solid-state image pickup device of an all-pixel readout method. However, the above-described defective pixel correction apparatus has a problem that the resolution of image information is reduced because two pixels are added in the horizontal direction to obtain a luminance signal and then defective pixel detection and correction are performed. Was.
[0007]
The present invention has been made to solve the conventional problems, and has as its object to provide a defective pixel correction device for a solid-state imaging device capable of obtaining a high-quality image.
[0008]
[Means for Solving the Problems]
The defective pixel correction device for a solid-state imaging device according to the present invention has a unique pixel signal among all-pixel readout solid-state imaging devices having a checkerboard color filter and capable of outputting all pixels independently and sequentially. A defective pixel detecting means for detecting a defective pixel which outputs a signal of an appropriate level, and a defective pixel correcting means for correcting a pixel determined as a defective pixel by the defective pixel detecting means are provided.
With this configuration, a defective pixel that outputs a signal of a peculiar level is detected from among the output signals of the solid-state imaging device of the all-pixel readout method, and a pixel determined as a defective pixel is corrected.
[0009]
Further, the defective pixel correction device for a solid-state imaging device according to the present invention, the defective pixel detection means, the determination target pixel to be determined whether there is a defect, and the same color of the determination target pixel other than the determination target pixel It has a configuration for determining a defective pixel based on the signal level of a pixel having a color filter.
With this configuration, the signal level of a defective pixel can be detected without erroneous detection even in a chromatic subject.
[0010]
Further, in the defective pixel correction device for a solid-state imaging device according to the present invention, the defective pixel detecting unit determines whether or not there is a defect, and the determination target pixel and the determination target pixel other than the determination target pixel have the same color as the determination target pixel. A configuration in which a pixel to be determined is determined to be a defective pixel when a signal level difference between the pixel to be determined and a pixel having a color filter of the same color is equal to or larger than a predetermined threshold value by comparing with a pixel having a color filter. Have.
With this configuration, it is possible to accurately determine whether the determination target pixel is a defective pixel.
[0011]
Further, the defective pixel correction device for a solid-state imaging device according to the present invention is configured such that the threshold value of the defective pixel detection means is optimally set according to the AGC gain of the AGC circuit provided at the input stage of the fixed imaging device. Have.
With this configuration, defective pixel correction can be performed when the AGC gain is large and the effect of image quality degradation due to defective pixels is large, and when the AGC gain is small and the effect of image quality degradation due to defective pixels is small, the operation of defective pixel correction can be stopped. .
[0012]
Furthermore, in the defective pixel correction device for a solid-state imaging device according to the present invention, when the defective pixel correction unit determines that the determination target pixel is a defective pixel, the defective pixel correction unit determines that the determination target pixel has no defect adjacent to the determination target pixel. It has a configuration to replace pixels with color filters.
With this configuration, a high-quality image that is not affected by defective pixels can be obtained.
[0013]
Further, the defective pixel correction device for a solid-state imaging device according to the present invention has a configuration in which detection of defective pixel means and correction of defective pixel correction means are simultaneously performed on pixels in a plurality of vertical lines.
With this configuration, all pixel signals input to the subsequent luminance signal processing circuit and color signal processing circuit have corrected defective pixels.
[0014]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, embodiments of the present invention will be described with reference to the drawings. The device for correcting a defective pixel of a solid-state image sensor in FIG. The solid-state imaging device 2 is an all-pixel readout solid-state imaging device 2 having a primary color checkerboard color filter and capable of outputting all pixels independently and sequentially. The solid-state imaging device 2 is connected to a pre-processing unit 3 including a CDS (Correlated Double Sampling) circuit and an AGC (Automatic Gain control) circuit. The CDS circuit is a correlated double sampling circuit for removing image noise, and the AGC circuit is a circuit for automatically adjusting the gain so that the output becomes constant when the level of the input signal fluctuates. It is.
[0015]
An A / D converter 4 for converting an analog signal into a digital signal is provided at a stage subsequent to the pre-processing unit 3, and thereafter, a white balance adjustment circuit (WB) 5 for adjusting a white balance is provided. The white balance adjustment circuit (WB) 5 is connected to the defective pixel detection means 7 and has three stages of 1H delay circuits 6 for delaying signals by one horizontal line, and outputs (signal b, signal c and the signal d) are connected to the defective pixel detecting means 7 at the subsequent stage. The output from the defective pixel detecting means 7 for determining whether or not the pixel is a defective pixel is connected to a defective pixel correcting means 8 and thereafter to a luminance signal processing circuit 9 and a chrominance signal processing circuit 10. The signal subjected to the defect correction is subjected to various kinds of signal processing in a luminance signal processing circuit 9 and a color signal processing circuit 10, and is output as a luminance signal and a color signal. The microcomputer 11 controls the operation of the entire camera such as exposure control such as threshold value and AGC gain used in the defective pixel correction device of the present invention, white balance control and the like.
[0016]
The operation of the defective pixel correction device for the solid-state imaging device 2 configured as described above will be described. First, incident light from a subject is focused on the imaging surface of the solid-state imaging device 2 by the lens 1 and is converted into an electric analog signal. The signal output from the solid-state imaging device 2 is subjected to noise removal and gain adjustment by the CDS circuit and the AGC circuit of the preprocessing unit 3 and then converted to a digital signal by the A / D converter 4. This signal is a signal a that is directly input to the defective pixel detecting means 7 after white balance adjustment is performed by the white balance adjustment circuit 5 and a signal delayed by one horizontal line through the 1H delay circuit 6 of one stage. b, a signal c delayed by one horizontal line through the 1H delay circuit 6 of the further stage, and a signal d delayed by one horizontal line through the 1H delay circuit 6 of the further stage are sent to the defective pixel detecting means 7. Sent. The defective pixel detecting means 7 detects a defective pixel based on the output signal from the solid-state imaging device 2, and the defective pixel correcting means 8 corrects the defective pixel detected by the defective pixel detecting means 7.
[0017]
Assuming a solid-state imaging device 2 of the all-pixel readout type having a primary color checkerboard color filter array as shown in FIG. 2, the output video signal of each line is G (green) → B ( A line repeating blue) → G (green) → B (blue) and a line repeating R (red) → G (green) → R (red) → G (green) are output alternately. The read video signal is composed of three lines of signals (signal b, signal c, and signal d) sequentially delayed by three 1H delay circuits 6 and four lines of an undelayed signal (signal a). At the same time, it is input to the defective pixel detection means 7. The states of the signals a, b, c and d in FIG. 1 are shown in FIGS. 4 (a), (b), (c) and (d), respectively. In the defective pixel detecting means 7, as shown in FIG. ο, π, ρ, τ, and υ are extracted, and it is determined whether or not the determination target pixels γ, θ, ν, and σ are defective pixels.
[0018]
Hereinafter, a determination method by the defective pixel detection means 7 will be described. FIG. 6 is a diagram for explaining the correspondence between the determination target pixel and the area of the peripheral pixel to be compared when the determination target pixel includes a green color filter pixel. When an image of a subject having a small change in luminance is taken, the correlation between adjacent same-color filter pixels is high and the signal level difference between them is small. On the other hand, in a defective pixel portion, an image defect in which a constant bias voltage is always added to an imaging output corresponding to the amount of incident light due to a low correlation with peripheral pixels appears as a high luminance point on an output screen. (Hereinafter, it is called a white defective pixel.) Conversely, it can be said that the one with low photoelectric sensitivity has low correlation with the surroundings as a low luminance point (hereinafter, called a black defective pixel). According to the present invention, a defective pixel is detected by calculating a signal level difference between a determination pixel and a peripheral pixel based on the prominence.
[0019]
FIG. 6A shows an area of a peripheral pixel to be compared when a G (green) pixel at a position corresponding to the c pixel in FIG. 5 is determined as a defective pixel. When the defective pixel is at a higher signal level than the surroundings, the white defective pixel determination is as follows. The symbols α to 用 い used in the equations and the description of the equations are symbols indicating the positions of the pixels shown in FIG. 5, and when used in the equations, indicate the signal level value of each pixel.
Judgment of white defective pixel | γ-max (α, ε, η, ι) | ≧ thw
That is, the difference between the signal level of the maximum value of the four surrounding pixels (α, ε, η, ι) and the signal level of the determination target pixel γ is calculated. Since it is large, it is determined to be a defective pixel. If it is determined that the pixel is a white defective pixel, the defective pixel correcting unit 8 replaces the value of the defective pixel γ with max (α, ε, η, ι), which is the maximum value of the four peripheral pixels, and outputs a luminance signal processing circuit 9 and the color signal processing circuit 10. If it is not determined that the pixel is a white defective pixel, the signal of the pixel γ is output as it is to the subsequent stage.
[0020]
Conversely, when the defective pixel is at a lower signal level than the surrounding pixels, the black defective pixel determination is as follows.
Black defective pixel determination | γ-min (α, ε, η, ι) | ≧ thb
That is, the difference between the minimum value of the four surrounding pixels (α, ε, η, ι) and the determination target pixel γ is determined. If the difference is equal to or larger than the predetermined threshold thb, the determination target pixel is prominently small. Is determined. When it is determined that the pixel is a black defective pixel, the defective pixel correcting unit 8 replaces the value of the defective pixel γ with min (α, ε, η, ι), which is the minimum value of the four peripheral pixels, and outputs the luminance signal processing circuit of the subsequent stage. 9 and the color signal processing circuit 10. If the pixel is not determined to be a black defective pixel, the signal of the pixel γ is output to the subsequent stage as it is. The above-described determination of a white defective pixel and determination of a black defective pixel are performed simultaneously.
[0021]
Similarly, as shown in FIG. 6B, the determination formula when the G (green) determination target pixel is at θ is as follows.
Determination of white defective pixel | θ-max (β, δ, ζ, κ, μ, ξ) | ≧ thw
If the pixel θ is determined to be a defective pixel, the defective pixel correction means 8 replaces the value of the defective pixel θ with max (β, δ, ζ, κ, μ, ξ), which is the maximum value of the six peripheral pixels, and outputs the result.
Judgment of black defective pixel | θ-min (β, δ, ζ, κ, μ, ξ) | ≧ thb
If the pixel θ is determined to be a defective pixel, the defective pixel correction means 8 replaces the value of the defective pixel θ with min (β, δ, ζ, κ, μ, ξ), which is the minimum value of the six peripheral pixels, and outputs the result.
[0022]
Then, as shown in FIG. 6C, the determination formula when the G (green) determination target pixel is at ν is as follows.
Determination of white defective pixel | ν-max (η, ι, λ, ο, ρ, τ) | ≧ thw
When the pixel ν is determined to be a defective pixel, the defective pixel correction unit 8 replaces the value of the defective pixel ν with max (η, ι, λ, o, ρ, τ), which is the maximum value of the six peripheral pixels, and outputs the result.
Judgment of black defective pixel | ν-min (η, ι, λ, ο, ρ, τ) | ≧ thb
When the pixel ν is determined to be a defective pixel, the defective pixel correction unit 8 replaces the value of the defective pixel ν with min (η, ι, λ, o, ρ, τ), which is the minimum value of the six peripheral pixels, and outputs the result.
[0023]
Further, as shown in FIG. 6D, the determination formula when the G (green) determination target pixel is at σ is as follows.
Judgment of white defective pixel | σ-max (μ, ξ, π, υ) | ≧ thw
If the pixel σ is determined to be a defective pixel, the defective pixel correction means 8 replaces the value of the defective pixel σ with the maximum value of the four surrounding pixels, max (μ, ξ, π, υ), and outputs it.
Judgment of black defective pixel | σ-min (μ, ξ, π, υ) | ≧ thb
If the pixel σ is determined to be a defective pixel, the defective pixel correction means 8 replaces the value of the defective pixel σ with min (μ, ξ, π, υ), which is the minimum value of the four surrounding pixels, and outputs the result.
[0024]
Here, thw and thb are threshold values for judging the protrusion amount, and are values that are appropriately set by the microcomputer 11 according to the AGC gain amount. With this configuration, defective pixel correction can be performed when the AGC gain is large and the effect of image quality degradation due to defective pixels is large, and when the AGC gain is small and the effect of image quality degradation due to defective pixels is small, the operation of defective pixel correction can be stopped. . For example, a white defect pixel is a pixel in which a constant bias voltage is always added to an imaging output corresponding to the amount of incident light. In a state where the amount of incident light from the subject decreases and the AGC gain increases, the amount of protrusion from peripheral pixels is It will be bigger. 9A and 9B show how the signal level of a white defective pixel changes depending on the AGC gain. FIG. 9A shows the signal level of a defective pixel when the AGC gain is small, and FIG. This is the defective pixel level when the number of pixels increases. Accordingly, when the subject condition in which the AGC gain is large is set, the threshold value of the protrusion amount is also increased, so that the defective pixel detection and correction can always be stably performed.
[0025]
The case where the determination target pixel is a G (green) color filter pixel has been described above, but the determination and correction of a defective pixel can be similarly performed for pixels of other colors. FIG. 7 is an explanatory diagram showing a determination target pixel and a peripheral pixel area when there is an R (red) color filter pixel as a determination target pixel, and FIG.
[0026]
In the present invention, the above-described defective pixel detection and correction may be simultaneously performed on the determination target pixels on a plurality of lines in the vertical direction as shown in FIG. In this case, all the pixel signals input to the subsequent luminance signal processing circuit and color signal processing circuit are obtained by correcting the defective pixels.
[0027]
As described above, the defective pixel correction device for the solid-state imaging device 2 of the present embodiment detects a defective pixel that outputs a signal of a peculiar level from among the output signals of the solid-state imaging device 2 that drives all the pixels to read out. Pixels determined to be pixels are corrected. Therefore, it is possible to detect and correct defective pixels without a memory for storing defective pixel data.
[0028]
Further, the defective pixel detection means 7 determines whether or not there is a defect based on the signal level of a pixel to be determined and a signal level of a pixel having a color filter of the same color as the determination target pixel other than the determination target pixel. Since the defective pixel is determined, the signal level of the defective pixel can be detected without erroneous detection even in a chromatic color subject. Note that a pixel having a color filter of the same color as the determination target pixel other than the determination target pixel may be a pixel adjacent to the determination target pixel or a peripheral pixel several pixels away, and so on.
[0029]
Further, the defective pixel detection means 7 compares the pixel to be determined as a target for determining whether or not there is a defect with a pixel having a color filter of the same color as the pixel to be determined other than the pixel to be determined. When the difference in signal level between the target pixel and a pixel having the same color filter is equal to or greater than a predetermined threshold, the determination target pixel is determined to be a defective pixel. The judgment can be made accurately.
[0030]
Further, when the pixel to be determined is determined to be a defective pixel, the defective pixel correction unit 8 replaces the pixel to be determined with a pixel adjacent to the pixel to be determined and having a color filter having the same color that has no defect. A high-quality image that is not affected by defective pixels can be obtained.
[0031]
【The invention's effect】
According to the present invention, a signal of a specific level is output among the output signals of the solid-state imaging device of the all-pixel readout method which has a checkerboard color filter of primary colors and can output all pixels independently and sequentially. By providing a defective pixel detecting unit for detecting a defective pixel, and a defective pixel correcting unit for correcting a pixel determined as a defective pixel by the defective pixel detecting unit, a defective pixel correction of a solid-state imaging device that can obtain a high-quality image is provided. An apparatus can be provided.
[Brief description of the drawings]
FIG. 1 is a configuration diagram of an apparatus for correcting a defective pixel of a solid-state imaging device according to an embodiment of the present invention. FIG. 2 is an explanatory diagram showing an example of a color filter array of the solid-state imaging device. FIG. 4 is an explanatory diagram showing signals of various parts in FIG. 1. FIG. 5 is an explanatory diagram showing an area for detecting a defective pixel. FIG. 6 is an explanatory diagram showing a defective pixel detection area of a G pixel. 7 is an explanatory diagram showing a defective pixel detection area of an R pixel. FIG. 8 is an explanatory diagram showing a defective pixel detection area of a B pixel. FIG. 9 is an explanatory diagram showing a defective pixel value when an AGC gain is increased. Configuration diagram of conventional defective pixel correction device for solid-state imaging device
1 lens 2 solid-state imaging device 3 pre-processing unit (CDS, AGC)
4 A / D converter 5 White balance adjustment circuit (WB)
6 1H delay circuit 7 defective pixel detecting means 8 defective pixel correcting means 9 luminance signal processing circuit 10 color signal processing circuit 11 microcomputer 12 low-pass filter (LPF)

Claims (6)

A defect that detects a defective pixel that outputs a signal of a peculiar level among output signals of a solid-state imaging device of an all-pixel readout type that has a primary color checkerboard color filter and can output all pixels independently and sequentially. A defective pixel correction device for a solid-state imaging device, comprising: a pixel detection unit; and a defective pixel correction unit that corrects a pixel determined as a defective pixel by the defective pixel detection unit. The defective pixel detection means is based on a signal level of a pixel to be determined as to whether or not there is a defect and a signal level of a pixel having the same color filter as the pixel to be determined other than the pixel to be determined. The defective pixel correction device for a solid-state image sensor according to claim 1, wherein the defective pixel is determined by performing the determination. The defective pixel detection unit performs a comparison between a determination target pixel that is a target for determining whether there is a defect and a pixel having the same color filter as the determination target pixel other than the determination target pixel, The method according to claim 1, wherein the determination target pixel is determined as a defective pixel when a difference in signal level between the determination target pixel and the pixel having the same color filter is equal to or greater than a predetermined threshold value. Defective pixel correction device for solid-state imaging device. 4. The defect of the solid-state imaging device according to claim 3, wherein a threshold value of the defective pixel detection unit is set optimally according to an AGC gain of an AGC circuit provided in an input stage of the fixed imaging device. Pixel correction device. The defective pixel correcting means, when the determination target pixel is determined to be the defective pixel, the determination target pixel to a pixel having the same color and having the same color and having no defect adjacent to the determination target pixel The defective pixel correction device for a solid-state imaging device according to claim 2, wherein the defective pixel correction device is replaced. The solid-state imaging device according to claim 1, wherein the detection of the defective pixel unit and the correction of the defective pixel correction unit are performed simultaneously on pixels in a plurality of vertical lines. Defect correction device.

Images