KR20070069873A - Apparatus and method for improving black signal in cmos image sensor - Google Patents

Abstract

An image sensor for improving a black signal and a black signal improving method of the image sensor are provided to remove a black line formed on a light-shielded region around an active cell and perform ABLC(Auto Black Level Compensation) by using data obtained from pixels to calculate a correct dark current value. A CMOS(Complementary Metal Oxide Semiconductor) image sensor includes a unit pixel, an active pixel array, and an ADC(Analog-to-Digital Conversion) block. The unit pixel includes a transfer transistor for transferring photocharges generated in a photo-diode to a sensing node in response to a first control signal and a reset transistor for discharging charges stored at the sensing node in response to a second control signal. The active pixel array includes a plurality of unit pixels. The CMOS image sensor includes a black line in the active pixel array in order to calculate a dark current.

Description

Black signal improvement image sensor and its improvement {apparatus and method for improving black signal in CMOS image sensor}

1A is a conventional Auto black level compensation (ABLC) block diagram.

1B is a conventional ABLC processing flow diagram.

2 is a structural diagram of a CMOS image sensor according to an embodiment of the prior art;

3 is a structural diagram of a CMOS image sensor according to another embodiment of the prior art;

4 is a unit pixel circuit diagram of a typical CMOS image sensor.

5 is an ABLC block diagram in accordance with the present invention.

6 is an ABLC processing flow chart in accordance with the present invention.

* Explanation of symbols for the main parts of the drawings

30: pixel array unit 33: analog line buffer unit

32a: Rod Decoder 32b: Column Decoder

Tx: transfer transistor Rx: reset transistor

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image sensor, and more particularly, to an image sensor and a method for improving dark current by minimizing the influence of a black signal (dark current) and obtaining a good image by utilizing an active cell of a pixel array unit as part of a black line. It is about.

In general, the field of use of the image sensor is a variety of still cameras, PC cameras, medical, toys, portable terminals. The image sensor generates black signals with different values in different environments, causing the output screen to become cloudy overall. Therefore, to get a good image, you have to solve the black signal that converts according to the environment.

CMOS image sensors have a black signal by default. The image output by this black signal looks whitish. In this case, the black signal refers to the signal code level that is output when no light enters the sensor. Theoretically, the code value should be 0 when no light is applied. However, in reality, even when no light is applied, data output by the basic noise of a pixel or an analog channel has a rather high code value, not 0 code. It usually has a value of about 5 codes, but in some cases, it may have a high black signal value of 10 codes or more.

In order to prevent the above black signal, an algorithm called auto black level compensation (ABLC) is used. This algorithm calculates the average of the output data in a state in which light is not applied at all by covering the two lines of the upper and lower lines of the active pixel 1 with the methyl (hereinafter referred to as the black line 2) as shown in FIG. It is. Since no light is applied at all, the theoretical output code value is 0, but when the actual average is calculated, a code value of about 5 codes is output. It also causes the output code to be zero. The ADC block 3 shown in the figure performs an operation of converting the received analog data into a digital value, and the digital block 4 performs a function of determining an offset value in the digitally converted value. Will be explained.

Referring to FIG. 1B, which illustrates the process of processing the black line of FIG. 1A, first, the overall average of the black lines is calculated, and the difference between the initial value of offsetDAC and the initial value of the offsetDAC is initially obtained. If it comes out, put-code value to offsetDAC and if it is lower than initial value, put + code value to offset DAC to match 0 code which is the theoretical code when light is not applied.

Looking specifically at the method of configuring each ADC data of the active cell, first, to read the data from the black line (11), to form a digital value in the ADC block (12), the Offset DAC (initial) value and the black line The sum of the data will form the ADC data. The digital block resets the OffsetDAC value by calculating the difference between the initial Offset value and the ADC value (13), and the reset OffsetDAC value is reflected in the active cell data in the ADC block.

Meanwhile, US Patent Publication No. 2003 / 0184666A1 discloses a method of adjusting the timing of an updated offset value, which will be described in more detail with reference to FIG. 2.

First, the first pixel group 20 which detects information about an image coming from the outside by arranging the pixels N horizontally and vertically M (N, M are natural numbers) to maximize the response to light. That is, the core pixel array unit and the second pixel group 21 and the light blocking are arranged on one side and the other side in the column direction of the first pixel group 20 to calculate an offset value of the black level with respect to the constituent pixels. The pixel array unit 30 including the third pixel group 22 for changing the offset value of the second pixel group 30 according to the offset value of the black level to remove the offset change caused by the black level. It is comprised with the offset adjustment part 31.

In detail, the analog line buffer unit 33 receives and buffers a signal from the pixel array unit 30, an output of the analog line buffer unit 33, that is, an amplifier 34 for amplifying pixel data and an ADC ( 35 and a rod decoder 32a for controlling the rows and columns of the pixel array unit 30 and the column decoder 32b, respectively, and the offset adjustment unit 31 includes an offset value of the calculated black level and an initial setting. A timing controller 31a for applying to the timing of the updated ADC offset for each of the colors R (red), G (green), and B (blue) according to the difference between the offset values (Initial offset), And an adder 31b for adding the updated offset value provided by the timing controller 31a and the analog pixel data from the first pixel group 20 to compensate for the analog data. The analog line buffer unit 33 serves to sense and store the voltage of the selected one row of pixels, and the amplifying unit 34, for example, the PGA, amplifies the pixel voltage stored in the analog line buffer unit 33 when it is small. The analog data passed through the amplifier 34 can adjust the gain of RGB for color interpolation and color correction, and is converted into a digital value through the ADC 35 and output.

However, there is a problem that it is impossible to calculate an accurate dark current value because the conventional timing adjustment also refers only to a black line of a specific region.

Meanwhile, referring to US Patent Publication No. 2005 / 0024502A1, a method of considering an offset through a black line in analog amplification before converting a pixel value to an ADC is disclosed. Specifically, referring to FIG. 3, when the output of the ADC converter and the reference signal (not shown) from the black line are input to generate a correction (error) signal in the digital block, the computing means and The DAC generates an offset voltage that adjusts the gain of the analog amplifier.

However, since the black line does not accurately represent the dark current value used in the image pixel part according to the number and position of the reference pixel, there is a problem in that the accurate dark current value cannot be calculated. Moreover, there are various environments that do not fit in this way. A representative case is for temperature.

As the temperature changes, the existing ABLC algorithm cannot remove the black signal properly. If the temperature of the chip rises above 40 ℃, the screen will deteriorate. Screen degradation refers to a phenomenon in which the color of an image loses its original color characteristics and has a different color. Turning off the ABLC function when screen degradation occurs eliminates screen degradation and preserves the original color. Therefore, if the temperature of the chip rises, screen degradation occurs due to the influence of ABLC.

In order to solve the above problem, the present invention removes black lines formed in a region where light is blocked around an active cell, and performs ABLC using data obtained from a pixel unit, thereby calculating a more accurate dark current value. It is an object of the present invention to provide an image sensor and a method for improving dark current.

In addition, another object of the present invention is to provide an image sensor and a method for improving dark current, which can provide an improved image by calculating a dark current using an active cell.

The present invention for achieving the above object, the sensing node in response to the transfer transistor (Tx) and the second control signal (Reset) for transporting the photocharge generated in the photodiode in response to the first control signal to the sensing node. A CMOS image sensor having a unit pixel having a reset transistor (Rx) for discharging charge stored in an active pixel array, an active pixel array having a plurality of unit pixels formed therein, and an ADC block, wherein the active is calculated to calculate a dark current. Some lines in the pixel array include black lines.

In addition, the present invention is stored in the sensing node in response to the transfer transistor (Tx) and the second control signal (Reset) for transporting the photocharge generated in the photodiode in response to the first control signal to the sensing node. In a method of improving a black signal of a CMOS image sensor having a unit pixel having a reset transistor Rx for discharging charge, an active pixel array having a plurality of unit pixels formed therein, and an ADC block, a separate independent line is provided. Without using, providing at least one black line to some lines in the active pixel array.

Hereinafter, the most preferred embodiments of the present invention will be described with reference to the accompanying drawings so that those skilled in the art may easily implement the technical idea of the present invention. 4 is a diagram illustrating a pixel structure of a general CMOS image sensor, FIG. 5 is a detailed configuration diagram of an image sensor performing an ABLC algorithm according to the present invention, and FIG. 6 is a flowchart illustrating an ABLC algorithm according to the present invention. to be.

First, the method proposed in the present invention has a key point in using the data of the active area that actually operates the line itself that performs the ABLC. The main reason for screen degradation is that the black signal value calculated through the ABLC line eventually differs from the actual value used. When the temperature rises, the difference between the value of the active region and the value of the ABLC line is basically due to structural differences. The existing ABLC line is covered with metal. The actual active pixel is not covered by any material. As the temperature rises, the ABLC line will output higher data, and the active pixel block that produces the actual image will require a lower black signal than the ABLC line.

Due to such basic structural characteristics, screen degradation due to temperature is difficult to avoid, and in order to obtain a black signal required by an active pixel, it is most accurate to obtain data based on the active pixel. It is one of the most important features of the present invention that the black signal code value is obtained by obtaining data of the active pixel region instead of using an ABLC line covered with metal.

As described above, in order to obtain an active black signal in any environment, it is most accurate to obtain data from an active pixel. First, referring to FIG. 4, the pixel structure of the CMOS image sensor includes the low voltage photodiode 100 and the photocharge generated in the photodiode 100 in response to the first control signal to the sensing node A. FIG. A transfer transistor Tx for transport, a reset transistor Rx for discharging charge stored in the sensing node A in response to the second control signal Reset for the next signal detection, and a source follower A drive transistor NM3 serving as a source follower and a select transistor NM4 for addressing switching can be included in response to the third control signal Select, and the load transistor NM5 bias voltage Vgg is included. It is driven by.

In order to obtain a black signal, the average data of the pixels to which no light is applied should be obtained. When adjusting this to the integration time, a value of 0 should be given, but since the integration time is basically 10 or more, You can't get data with the light blocked. Therefore, in order to obtain completely blocked data, it is necessary to adjust the pixel controller to obtain the data.

To get the light completely blocked, open the transfer transistor (Tx) and reset transistor (Rx) in the pixel controller to keep the pixel level at 0, and then immediately close the Tx while closing the Tx and Rx when reading the pixel line data. You must open it once to read the data. This allows you to obtain data that is almost similar to the blocked light.

Referring to FIG. 5, the active line used is divided into six equal parts of all active pixels so as to have all characteristics among the pixels constituting the actual screen, and the remaining four boundary lines except the boundary line are set as active lines to obtain data. Do it. If the height is 1200, the row address for the black signal is 200, 400, 800, 1000. Next, when the line is read while the transmission transistor (Tx) and reset transistor (Rx) are kept high to implement the state where no light is applied, the transmission transistor (Tx) and the reset transistor (Rx) are simultaneously This is done by reading the data while making it low and making the transfer transistor (Tx) high on the next clock. This is an operation that does not apply light to the pixel at all.

6 is a flowchart illustrating an ABLC algorithm according to an embodiment of the present invention. Looking at the specific operation in detail as follows.

First, an average data value for the ABLC 4 lines (address = 200, 400, 800, 1000) set in the active pixel block is calculated in a state where 7f or 3f and an initial value are given. At this time, when the transfer transistor Tx and the reset transistor Rx are turned on (61), when the corresponding line is read, the transfer transistor Tx and the reset transistor Rx are turned off at the same time. The data value of each line can be read by turning on the transfer transistor Tx at the next clock (62). Of course, it will be apparent to those skilled in the art that the controller should provide each control signal to have the proper timing of the transfer transistor Tx and reset transistor Rx.

Subsequently, the ADC block calculates the ADC data value by adding the data value of the active cell and the initial value OffsetDAC (63). In the digital block, the offset DAC value is newly set by calculating the difference between the initial value and the calculated average value. In operation 64, the ADC block calculates a new ADC data value by adding the offset DAC value calculated from the active pixel and the pixel data value.

Although the technical spirit of the present invention has been described in detail according to the above preferred embodiment, it should be noted that the above embodiment is for the purpose of description and not of limitation. In addition, those skilled in the art will understand that various embodiments are possible within the scope of the technical idea of the present invention.

According to the present invention as described above, a black signal suitable for an active pixel can be obtained in any environment, and thus, a phenomenon such as screen degradation can be prevented and data close to reality can be obtained. In addition, since there is no need for an additional line, as the existing black line is eliminated above and below the pixel array, the size of the entire chip can be reduced.

In addition, the present invention can obtain a black signal value approximating the actual value in various environments by calculating the black signal code from the active pixel itself, and thus can perform an effective auto black level compensation (ABLC) operation.

Claims (5)

Reset for discharging the charge stored in the sensing node in response to the second control signal (Reset) and the transfer transistor (Tx) for transporting the photocharge generated in the photodiode in response to the first control signal to the sensing node A CMOS image sensor having a unit pixel having a transistor Rx, an active pixel array in which a plurality of unit pixels are formed, and an ADC block And a black line on some lines in the active pixel array to produce a dark current. The method of claim 1, A controller for reading the blackline, And the controller turns on the transfer transistor (Tx) and the reset transistor (Rx) when addressing the corresponding black line. Reset for discharging the charge stored in the sensing node in response to the second control signal (Reset) and the transfer transistor (Tx) for transporting the photocharge generated in the photodiode in response to the first control signal to the sensing node A method of improving a black signal of a CMOS image sensor having a unit pixel including a transistor Rx, an active pixel array in which a plurality of unit pixels are formed, and an ADC block, Providing at least one black line to some lines in the active pixel array without using separate independent lines. The method of claim 3, wherein The method, Calculating an average data value for a plurality of black lines preset in the active pixel block; Calculating an ADC data value by adding a data value of the active cell and a given initial value in an ADC converter; Setting a new offset DAC value by obtaining a difference between the initial value and the calculated average value in the digital block; And In the ADC block, the method further comprises calculating a new ADC data value by adding an offset DAC value and a pixel data value calculated in the active pixel. The method of claim 4, wherein Computing the average data value, In a state in which the transfer transistor Tx and the reset transistor Rx are turned on (61), when the corresponding line is read, the transfer transistor Tx and the reset transistor Rx are simultaneously turned off; And Turning on the transfer transistor (Tx) at the next clock.

Images