阅读说明：本技术 图像噪声抑制装置及其方法 (Image noise suppression device and method thereof ) 是由 邱仲毅 黃文聰 于 2020-05-11 设计创作，主要内容包括：一种图像噪声抑制装置及方法。该方法包括：接收原始图像,原始图像包括多个图像像素,各个图像像素具有图像像素值及固定噪声值,固定噪声值包括小数；依据小数的比特数、图像像素值及固定噪声值,获得位移后图像像素值、位移后固定噪声值及随机数；依据位移后图像像素值、位移后固定噪声值、随机数及抑制方程式,获得位移后噪声抑制图像像素值；依据小数的比特数及位移后噪声抑制图像像素值,获得噪声抑制图像像素值；以及,依据噪声抑制图像像素值,获得噪声抑制图像。(An image noise suppression device and method. The method comprises the following steps: receiving an original image, the original image comprising a plurality of image pixels, each image pixel having an image pixel value and a fixed noise value, the fixed noise value comprising a decimal; obtaining a displaced image pixel value, a displaced fixed noise value and a random number according to the bit number of the decimal, the image pixel value and the fixed noise value; obtaining a post-displacement noise suppression image pixel value according to the post-displacement image pixel value, the post-displacement fixed noise value, the random number and a suppression equation; obtaining a noise suppression image pixel value according to the bit number of the decimal and the post-displacement noise suppression image pixel value; and obtaining a noise-suppressed image according to the noise-suppressed image pixel value.)

1. An image noise suppression method, comprising:

receiving an original image, the original image comprising a plurality of image pixels, each image pixel having an image pixel value and a fixed noise value, the fixed noise value comprising a decimal;

obtaining a displaced image pixel value, a displaced fixed noise value and a random number according to the bit number of the decimal, the image pixel value and the fixed noise value;

obtaining a post-displacement noise suppression image pixel value according to the post-displacement image pixel value, the post-displacement fixed noise value, the random number and a suppression equation, wherein the suppression equation is as follows:

O＝I-F2+R

wherein, O is the post-displacement noise suppressed image pixel value, I is the post-displacement image pixel value, F2 is the post-displacement fixed noise value, and R is the random number;

obtaining a noise suppression image pixel value according to the bit number of the decimal and the post-displacement noise suppression image pixel value; and

and obtaining a noise suppression image according to the noise suppression image pixel values.

2. The image noise suppression method according to claim 1, further comprising a parameter obtaining process for obtaining the shifted image pixel value, the shifted fixed noise value and the random number according to the number of bits of the decimal, the image pixel value and the fixed noise value, the parameter obtaining process comprising:

obtaining the random number according to the bit number of the decimal and a random function;

obtaining the pixel value of the image after the displacement according to the bit number of the decimal and the pixel value of the image; and

and obtaining the fixed noise value after the displacement according to the bit number of the decimal and the fixed noise value.

3. The image noise suppression method according to claim 2, wherein the random number is in a range of-2ⁿ<R<2ⁿR is the random number, and n is the number of bits of the decimal.

4. The image noise suppression method according to claim 1, wherein the random number is in a range of-16 < R <16 when the number of bits of the fractional number is 4.

5. The image noise suppression method of claim 1, further comprising an image capturing method for obtaining the received original image, the image capturing method comprising:

setting the amplification gain of an image capturing circuit to the maximum gain, and capturing a test image aiming at the scene, wherein the test image comprises a plurality of test pixels, the test pixels correspond to the image pixels, and each test pixel has the maximum fixed noise value;

setting the amplification gain of the image capturing circuit to a current gain, and capturing a current image for the scene;

obtaining the fixed noise value corresponding to each image pixel according to a noise adjustment equation:

F1＝(C/M)*F0

wherein F1 is the fixed noise value, C is the current gain, M is the maximum gain, and F0 is the maximum fixed noise value; and

and obtaining the original image according to the current image and the fixed noise value corresponding to each image pixel.

6. An image noise suppression apparatus comprising:

a processing circuit for obtaining a noise suppressed image according to an image noise suppression process, the image noise suppression process comprising:

receiving an original image, the original image comprising a plurality of image pixels, each image pixel having an image pixel value and a fixed noise value, the fixed noise value comprising a decimal;

obtaining a displaced image pixel value, a displaced fixed noise value and a random number according to the bit number of the decimal, the image pixel value and the fixed noise value;

obtaining a post-displacement noise suppression image pixel value according to the post-displacement image pixel value, the post-displacement fixed noise value, the random number and a suppression equation, wherein the suppression equation is as follows:

O＝I-F2+R

wherein, O is the post-displacement noise suppressed image pixel value, I is the post-displacement image pixel value, F2 is the post-displacement fixed noise value, and R is the random number;

obtaining a noise suppression image pixel value according to the bit number of the decimal and the post-displacement noise suppression image pixel value; and

and obtaining a noise suppression image according to the noise suppression image pixel values.

7. The image noise suppression device of claim 6, wherein the processing circuit is configured to obtain the shifted image pixel value, the shifted fixed noise value and the random number according to a parameter obtaining process, the parameter obtaining process including:

obtaining the random number according to the bit number of the decimal and a random function;

obtaining the pixel value of the image after the displacement according to the bit number of the decimal and the pixel value of the image; and

and obtaining the fixed noise value after the displacement according to the bit number of the decimal and the fixed noise value.

8. The image noise suppression device according to claim 7, wherein the random number is in a range of-2ⁿ<R<2ⁿR is the random number, and n is the number of bits of the decimal.

9. The image noise suppressing apparatus according to claim 6, wherein the random number is in a range of-16 < R <16 when the number of bits of the fractional number is 4.

10. The image noise suppression device of claim 6, further comprising an image capturing circuit for obtaining the received original image according to an image capturing process, the image capturing process comprising:

setting the amplification gain of the image capturing circuit to the maximum gain, and capturing a test image aiming at the scene, wherein the test image comprises a plurality of test pixels, the test pixels correspond to the image pixels, and each test pixel has the maximum fixed noise value;

setting the amplification gain of the image capturing circuit to a current gain, and capturing a current image for the scene;

obtaining the fixed noise value corresponding to each image pixel according to a noise adjustment equation:

F1＝(C/M)*F0

wherein F1 is the fixed noise value, C is the current gain, M is the maximum gain, and F0 is the maximum fixed noise value; and

and obtaining the original image according to the current image and the fixed noise value corresponding to each image pixel.

Technical Field

The present application relates to the field of image noise suppression, and in particular, to an image noise suppression apparatus and method thereof.

Background

An image sensor is a device that converts an optical signal into an electrical signal, and is also called a photosensitive device. In the process of capturing images of a scene by image sensors, the images finally generated by the image sensors are affected by various noises, such as crosstalk between light beams of various frequencies (e.g. red light, green light, blue light, infrared light), crosstalk between elements of the image sensors and light beams, or noises generated by the elements of the image sensors themselves or operations among each other. When the image sensor does not perform corresponding suppression on these different noises, the image generated by the image sensor is often distorted and erroneous with the scene.

Fixed Pattern Noise (Fix Pattern Noise) in an image, such as line-to-line brightness differences, is also one of the problems of image distortion. However, the conventional fixed noise suppression method still cannot effectively solve the problem of the fixed pattern noise, which causes the quality of the image to be seriously affected.

Disclosure of Invention

In view of the foregoing, the present application provides an image noise suppression device and a method thereof.

In accordance with some embodiments, an image noise suppression method includes: receiving an original image, the original image comprising a plurality of image pixels, each image pixel having an image pixel value and a fixed noise value, the fixed noise value comprising a decimal; obtaining a displaced image pixel value, a displaced fixed noise value and a random number according to the bit number of the decimal, the image pixel value and the fixed noise value; obtaining a post-displacement noise suppression image pixel value according to the post-displacement image pixel value, a post-displacement fixed noise value, a random number and a suppression equation, wherein the suppression equation is O-F2 + R, O is the post-displacement noise suppression image pixel value, I is the post-displacement image pixel value, F2 is the post-displacement fixed noise value, and R is the random number; obtaining a noise suppression image pixel value according to the bit number of the decimal and the post-displacement noise suppression image pixel value; and obtaining a noise-suppressed image according to the noise-suppressed image pixel value.

According to some embodiments, an image noise suppression apparatus includes a processing circuit configured to obtain a noise suppressed image according to an image noise suppression process. The image noise suppression process includes: receiving an original image, the original image comprising a plurality of image pixels, each image pixel having an image pixel value and a fixed noise value, the fixed noise value comprising a decimal; obtaining a displaced image pixel value, a displaced fixed noise value and a random number according to the bit number of the decimal, the image pixel value and the fixed noise value; obtaining a post-displacement noise suppression image pixel value according to the post-displacement image pixel value, a post-displacement fixed noise value, a random number and a suppression equation, wherein the suppression equation is O-F2 + R, O is the post-displacement noise suppression image pixel value, I is the post-displacement image pixel value, F2 is the post-displacement fixed noise value, and R is the random number; obtaining a noise suppression image pixel value according to the bit number of the decimal and the post-displacement noise suppression image pixel value; and obtaining a noise-suppressed image according to the noise-suppressed image pixel value.

In summary, according to some embodiments of the present disclosure, the image noise suppression apparatus and the method thereof can obtain each corresponding noise suppression image pixel value according to the fixed noise value and the corresponding random number of each image pixel in the original image, and then obtain the noise suppression image according to the noise suppression image pixel value. The image noise suppression device and the method thereof obtain a shifted image pixel value, a shifted fixed noise value and a random number according to the bit number of the decimal of the fixed noise value, the image pixel value and the fixed noise value, obtain a shifted noise suppression image pixel value according to the shifted image pixel value, the shifted fixed noise value, the random number and a suppression equation, and obtain a noise suppression image pixel value according to the bit number of the decimal of the fixed noise value and the shifted noise suppression image pixel value.

Drawings

Fig. 1 illustrates a schematic diagram of an image noise suppression apparatus according to some embodiments of the present application.

FIG. 2 illustrates a flow diagram of an image noise suppression method according to some embodiments of the present application.

FIG. 3 illustrates a schematic diagram of an original image in accordance with some embodiments of the present application.

FIG. 4 illustrates a flow diagram of a parameter acquisition method according to some embodiments of the present application.

Fig. 5 shows a schematic diagram of an image noise suppression device according to further embodiments of the present application.

FIG. 6 is a flow chart of an image capture method according to some embodiments of the present application.

FIG. 7 illustrates a schematic diagram of an original image according to further embodiments of the present application.

Detailed Description

The present application relates to image noise suppression. While several preferred modes of carrying out the application have been described in the specification, it is to be understood that the application can be carried out in numerous ways, and should not be limited to the specific embodiments described below or to the specific ways in which the features described below can be carried out. In other instances, well-known details will not be set forth or discussed in order to avoid obscuring the present application.

Fig. 1 illustrates a schematic diagram of an image noise suppression apparatus 10 according to some embodiments of the present application, and fig. 2 illustrates a flow chart of an image noise suppression method according to some embodiments of the present application. Referring to fig. 1 and fig. 2, in some embodiments, the image noise suppression apparatus 10 includes a processing circuit 100. The processing circuit 100 is used for obtaining a noise suppression image IM2 according to an image noise suppression process. Specifically, the processing circuit 100 performs an image noise suppression process to convert the original image IM1 into a noise-suppressed image IM 2. The image noise suppression process includes the steps of: an image receiving step (step S210); a parameter obtaining step (step S220); a noise suppression step (step S230); a displacement adjustment step (step S240); and a noise suppressed image obtaining step (step S250).

FIG. 3 illustrates a schematic diagram of an original image IM1 in accordance with some embodiments of the present application. Referring to fig. 1, fig. 2 and fig. 3, in some embodiments, the image receiving step (step S210 of fig. 2) includes: the original image IM1 is received, the original image IM1 includes a plurality of image pixels PX 1. Each image pixel PX1 has an image pixel value and a fixed noise value, which includes a decimal. Specifically, the fixed noise value received by the processing circuit 100 is a value including an integer and a decimal, the number of bits of the decimal (i.e. the precision of the number of bits after the decimal point is reached) can be determined by a user or a predetermined value in the processing circuit 100, and both the integer of the fixed noise value and the decimal of the fixed noise value can be represented by a decimal or a binary number. In other words, the fixed noise value received by the processing circuit 100 is represented, for example, but not limited to, in decimal or binary. For example, the fixed noise value of a decimal is "3.6", where the integer of the decimal is "3" and the decimal is "6". The fixed noise value of the binary is "0011.1010", wherein the integer of the binary is "0011", and the decimal of the binary is "1010", i.e., the integer and decimal of the binary are represented by 4 bits, but not limited thereto. In some embodiments, the units of the integer and the decimal of the binary may be different, for example, when the fixed noise value of the binary is "00000001.1000", the unit of the integer is "8 bits", and the unit of the decimal is "4 bits". When the fixed noise value of the binary is "01.1000", the unit of the integer is "2 bits", and the unit of the decimal is "4 bits". When the fixed noise value of the binary is "1.1000", the unit of the integer is "1 bit", and the unit of the decimal is "4 bits". When the fixed noise value of the binary is "1.10", the unit of the integer is "1 bit", and the unit of the decimal is "2 bit". In some embodiments, the image pixels PX1 are arranged in a two-dimensional array according to the first axis direction D1 and the second axis direction D2.

In some embodiments, the parameter obtaining step (step S220 of fig. 2) comprises: and obtaining the pixel value of the image after displacement, the fixed noise value after displacement and the random number according to the bit number of the decimal, the pixel value of the image and the fixed noise value. Specifically, the processing circuit 100 first obtains the number of bits of the decimal according to the decimal of the fixed noise value (which may be determined by a user or a predetermined value). Then, the processing circuit 100 obtains the shifted image pixel value, the shifted fixed noise value, and the random number according to the bit number of the decimal, the image pixel value, and the fixed noise value. In some embodiments, the processing circuit 100 obtains the number of bits of the fractional number in terms of units of the fractional number of the fixed noise value. For example, if the unit of the decimal is 4 bits, the number of bits of the decimal is "4". Or, when the unit of the decimal is 8 bits, the bit number of the decimal is "8".

FIG. 4 illustrates a flow diagram of a parameter acquisition method according to some embodiments of the present application. Referring to fig. 1, fig. 2 and fig. 4, in some embodiments, the processing circuit 100 is configured to obtain a parameter obtaining process (i.e., a parameter obtaining step, step S220 of fig. 2) according to a number of bits of a decimal, an image pixel value and a fixed noise value, and obtain a shifted image pixel value, a shifted fixed noise value and a random number. Specifically, the processing circuit 100 performs a parameter obtaining process to convert the number of decimal digits, the image pixel value, and the fixed noise value into a shifted image pixel value, a shifted fixed noise value, and a random number. The parameter obtaining process comprises the following steps: a random number obtaining step (step S222); a post-displacement image pixel value obtaining step (step S224); and a post-displacement fixed noise value obtaining step (step S226). Specifically, the order in which the processing circuit 100 executes steps S222, S224, and S226 is not limited to that shown in fig. 4, and the processing circuit 100 may execute steps S222, S224, and S226 in any order, or may execute any combination of steps S222, S224, and S226 simultaneously or non-simultaneously.

In some embodiments, the random number obtaining step (step S222 of fig. 4) includes: and obtaining a random number according to the bit number of the decimal and the random function. Specifically, in some embodiments, the processing circuit 100 obtains the random number according to the number of bits of the fractional number and a random function. Such as, but not limited to, Cyclic Redundancy Check (CRC), among others. In some embodiments, the random number ranges from-2ⁿ<R<2ⁿR is a random number, and n is the number of decimal bits. And in some embodiments the random number is an integer. In some embodiments, when the number of decimal bits is "4", the random number ranges from-16<R<16. That is, the random number is not more than 16, and not less than-16. For example, a decimal random number can be "9".

In some embodiments, the post-displacement image pixel value obtaining step (step S224 of fig. 4) includes: and obtaining the pixel value of the image after displacement according to the bit number of the decimal and the pixel value of the image. Specifically, the processing circuit 100 is configured to shift the decimal point of the binary image pixel value to the right by a number of bits of "decimal number of bits" to obtain the shifted image pixel value. The image pixel values include image pixel integer values and image pixel fractional values. For example, when the number of bits of the decimal is "4" and the binary image pixel value is "00100000.0000", that is, when the binary image pixel integer value is "00100000" and the binary image pixel decimal value is "0000", the binary shifted image pixel value is "001000000000". In decimal notation, when the decimal image pixel integer value is "32" and the decimal image pixel decimal value is "0", the decimal shifted image pixel is "512". In other words, the processing circuit 100 obtains the shifted image pixel value "001000000000" based on the decimal bit number "4" and the two-bit image pixel value "00100000.0000", which is equivalent to shifting the decimal point of the image pixel value by "4" bits to the right to obtain the shifted image pixel value.

In some embodiments, the post-displacement fixed noise value obtaining step (step S226 of fig. 4) includes: and obtaining a fixed noise value after the displacement according to the bit number of the decimal and the fixed noise value. Specifically, step S226 is similar to step S224, except that the processing circuit 100 is configured to shift the decimal point of the binary fixed noise value to the right by the number of bits of "decimal number of bits" to obtain the shifted fixed noise value. For example, when the number of bits of the decimal is "4" and the fixed noise value of the binary is "00000001.1001", that is, when the integer of the binary is "00000001" and the decimal of the binary is "1001", the fixed noise value after the shift of the binary is "000000011001". Expressed in decimal, when the integer of decimal is "1" and the decimal is "9", the fixed noise value after decimal displacement is "25". In other words, the processing circuit 100 obtains the post-shift fixed noise value "000000011001" according to the decimal bit number "4" and the binary fixed noise value "00000001.1001", which is equivalent to obtaining the post-shift fixed noise value by shifting the decimal point of the fixed noise value by "4" digits to the right.

In some embodiments, the noise suppression step (step S230 of fig. 2) comprises: and obtaining the pixel value of the post-displacement noise suppression image according to the pixel value of the post-displacement image, the post-displacement fixed noise value, the random number and the suppression equation. The suppression equation is O — I-F2+ R, where O is the post-displacement noise suppressed image pixel value, I is the post-displacement image pixel value, F2 is the post-displacement fixed noise value, and R is a random number. Specifically, after the processing circuit 100 obtains the post-displacement image pixel value, the post-displacement fixed noise value, and the random number, the post-displacement noise-suppressed image pixel value can be obtained through the suppression equation. It should be noted that, when the processing circuit 100 runs the suppression equation, the shifted image pixel value, the shifted fixed noise value and the random number are converted into the same carry form (for example, binary or decimal), and the shifted noise suppression image pixel value in the corresponding carry form is obtained. In some embodiments, for decimal, the image pixel I is "512" after shifting, the fixed noise value F2 is "25" after shifting, and the random number R is "9" after decimal. The processing circuit 100 obtains the corresponding post-displacement noise-suppressed image pixel value O as "496" according to the suppression equation. Similarly, taking the binary bit as an example, the image pixel I after the shift is "001000000000", the fixed noise value F2 after the shift is "000000011001", and the random number R is "000000001001". The processing circuit 100 obtains the corresponding post-displacement noise-suppressed image pixel value O of 000111110000 according to the suppression equation.

In some embodiments, the displacement adjusting step (step S240 of fig. 2) includes: and obtaining a noise suppression image pixel value according to the bit number of the decimal and the post-displacement noise suppression image pixel value. Specifically, the processing circuit 100 is configured to shift the decimal point of the binary bit shifted noise-suppressed image pixel value by a number of bits of "decimal number of bits" to the left to obtain the noise-suppressed image pixel value. For example, if the decimal bit number is "4" and the binary post-displacement noise-suppressed picture pixel value is "000111110000", the binary noise-suppressed picture pixel value is "00011111.0000". Expressed in decimal, when the noise suppressed picture pixel value after decimal displacement is "496", the noise suppressed picture pixel value in decimal is "31". In other words, the processing circuit 100 obtains the noise-suppressed image pixel value of "00011111.0000" based on the decimal bit number "4" and the binary shifted noise-suppressed image pixel value "000111110000", which corresponds to shifting the decimal point of the shifted noise-suppressed image pixel value by "4" bits to the left to obtain the noise-suppressed image pixel value.

Referring to fig. 1, fig. 2 and fig. 3, the noise-suppressed image obtaining step (step S250 of fig. 2) includes: the noise suppressed image IM2 is obtained from the noise suppressed image pixel values. Specifically, the noise suppression image IM2 includes a plurality of noise suppression image pixels. In other words, the arrangement of the noise suppressed image pixels in the noise suppressed image IM2 corresponds to the arrangement of the image pixels PX1 in the original image IM 1. Each noise-suppressed image pixel has a corresponding noise-suppressed image pixel value, where the noise-suppressed image pixel values correspond to the image pixel values in a one-to-one manner. Therefore, the processing circuit 100 can obtain the noise suppression image IM2 according to the noise suppression image pixel value. In some embodiments, the noise suppressed image pixels are arranged in a two-dimensional array according to a first axis D1 and a second axis D2.

Fig. 5 is a schematic diagram of an image noise suppression apparatus 10' according to some other embodiments of the present application, and fig. 6 is a flowchart of an image capturing method according to some embodiments of the present application. Referring to fig. 5 and fig. 6, in some embodiments, the image noise suppression apparatus 10' further includes an image capturing circuit 300, and the image capturing circuit 300 is coupled to the processing circuit 100, relative to the image noise suppression apparatus 10. The image capturing circuit 300 is used for obtaining the received original image IM1 according to the image capturing process. Specifically, the image capturing circuit 300 performs an image capturing process to capture the scene SC as the original image IM 1. Moreover, the image capturing circuit 300 transmits the original image IM1 to the processing circuit 100, so that the processing circuit 100 can receive the original image IM 1. The image noise suppression process includes the steps of: a maximum fixed noise value obtaining step (step S410); a current image obtaining step (step S420); a noise adjustment step (step S430); and an original image obtaining step (step S440).

FIG. 7 illustrates a schematic diagram of an original image according to further embodiments of the present application. Referring to fig. 5, fig. 6 and fig. 7, in some embodiments, the maximum fixed noise value obtaining step (step S410 of fig. 6) includes: the amplification gain of the image capturing circuit 300 is set to the maximum gain, and the test image IM0 is captured for the scene SC. The test image IM0 includes a plurality of test pixels PX0, the test pixels PX0 corresponding to the image pixels PX1, each test pixel PX0 having a maximum fixed noise value. Specifically, the amplification gain of the image capturing circuit 300 can be set between a maximum gain and a minimum gain, wherein the maximum gain is the maximum value of the amplification gain, and the minimum gain is the minimum value of the amplification gain. When the amplification gain of the image capturing circuit 300 is set to the maximum gain, the image captured by the image capturing circuit 300 for the scene SC is the test image IM 0. Each test pixel PX0 has a maximum fixed noise value, i.e., each test pixel PX0 has a corresponding maximum fixed noise value. Here, the test pixels PX0 in the test image IM0 correspond to the image pixels PX1 in the original image IM1 in a one-to-one manner, and the arrangement of the test pixels PM0 corresponds to the arrangement of the image pixels PX 1. In some embodiments, the test pixel PX0 has a "maximum fixed noise value" which is a "maximum row fixed noise value", i.e., each test pixel PX0 located in the same "row" has the same "maximum row fixed noise value".

In some embodiments, the current image obtaining step (step S420 of fig. 6) includes: the amplification gain of the image capture circuit 300 is set to the current gain, and the current image is captured for the scene SC. Specifically, the current gain is an amplification gain between a maximum gain and a minimum gain. When the amplification gain of the image capturing circuit 300 is set to the current gain, the image captured by the image capturing circuit 300 for the scene SC is the current image. Therefore, when the current gain is equal to the maximum gain, the current image is the test image IM 0.

In some embodiments, the image capture circuit 300 can dynamically adjust the current gain, such as but not limited to the following examples: the image capturing circuit 300 dynamically adjusts the current gain according to the background brightness of the scene SC (for example, when the background brightness of the scene SC is dark, the image capturing circuit 300 can increase the current gain so that the captured current image does not appear to be not bright enough due to the dark background brightness). When the image capturing circuit 300 reaches the maximum value or the minimum value according to the self-set auto-exposure, the image capturing circuit 300 dynamically adjusts the current gain to complement the brightness of the current image. The image capturing circuit 300 dynamically adjusts the current gain according to the adjustment instructions of other circuits of the image noise suppressing apparatus 10' (for example, when the other circuits are the processing circuit 100, the processing circuit 100 outputs different adjustment instructions according to different image capturing circuits 300, because the current images obtained by different image capturing circuits 300 have different fixed noise values).

In some embodiments, the noise adjusting step (step S430 of fig. 6) comprises: according to the noise adjustment equation, a fixed noise value corresponding to each image pixel PX1 is obtained. The noise adjustment equation is: f1 ═ C/M × F0, where F1 is the fixed noise value, C is the current gain, M is the maximum gain, and F0 is the maximum fixed noise value. Specifically, the fixed noise value corresponds to the current gain, and the fixed noise value changes according to the change of the current gain of the image capturing circuit 300, so that the image capturing circuit 300 can obtain the corresponding fixed noise value according to the fixed noise value, the current gain and the maximum gain corresponding to each image pixel PX 1. Since the maximum current gain is the maximum gain and the minimum current gain is the minimum gain, the value of "current gain divided by maximum gain C/M" is between 0 and 1. In some embodiments, each image pixel PX1 in the same original image IM1 corresponds to the same set of current gain and maximum gain, i.e., each image pixel PX1 in the same original image IM1 corresponds to the same "current gain divided by maximum gain C/M".

In some embodiments, the original image obtaining step (step S440 of fig. 6) includes: the original image IM1 is obtained according to the current image and the corresponding fixed noise value of each image pixel PX 1. Specifically, the current image is the original image IM1 without the fixed noise value, so the image capturing circuit 300 can obtain the original image IM1 according to the current image and the fixed noise value corresponding to each image pixel PX 1.

Referring to fig. 5 and fig. 7, in some embodiments, the image capturing circuit 300 is configured to obtain a maximum row fixed noise value corresponding to each test pixel PX0 according to a row fixed noise value obtaining process. Specifically, the maximum row fixed noise value is a maximum fixed noise value in units of "row", that is, the maximum fixed noise values of the test pixels PX0 in the same "row" are all the maximum row fixed noise values. The row fixed noise value obtaining process includes steps S510 to S550 (not shown), as follows:

in some embodiments, step S510 includes setting the amplification gain of the image capturing circuit 300 to the maximum gain, and capturing a plurality of test images IM0 for the scene SC. Specifically, the test image IM0 includes a plurality of rows C0, each row C0 being parallel to the second axis D2, each row C0 having a plurality of test pixels PX 0. In some embodiments, the background brightness of the scene SC used to obtain the plurality of test images IM0 is a low brightness, where the low brightness is a background brightness with an illuminance below 10 lux.

In some embodiments, step S520 includes averaging the sum of the test images IM0 according to an averaging equation to obtain an average test image. The average equation isWhere T is the average test image, Ti is the ith test image IM0, and N is the total number of test images IM0 summed. In particular, this step is used to eliminate general noise interference.

In some embodiments, step S530 includes obtaining a single-row average pixel value according to a single-row average pixel value obtaining formula, the test pixel values of the respective test pixels PX0, and the number of test pixels PX0 in each row C0. The single-line average pixel value is obtained by the formulaWhere Cx is the single-row average pixel value, T (x, y) is the test pixel value of test pixel PX0 at coordinate (x, y), and H is the number of test pixels PX0 in each row C0.

In some embodiments, step S540 includes obtaining the pixel values of the row neighboring regions of each row C0 according to the row neighboring region pixel value obtaining formula, the single-row average pixel value, and the width of the row neighboring region. The pixel value of the adjacent area of the line is obtained by the formulaWhere w is the width of the adjacent region of the row and Cx is the single row average pixel value. For example, when the width w of the row adjacent area is 2, the row adjacent area pixel value E5 of the "5" th row is (C3+ C4+ C5+ C6+ C7)/5, and C3 to C7 are the single-row average pixel values of the "3" th row to the "7" th row, respectively.

In some embodiments, step S550 includes obtaining the maximum row fixed noise value of each row C0 according to the maximum row fixed noise value obtaining formula, the single-row average pixel value, and the row adjacent area pixel value. The maximum line fixed noise value is obtained by the formula Fx-Cx-Ex, where Fx is the maximum line fixed noise value, Cx is the single-line average pixel value, and Ex is the line adjacent region pixel value.

In some embodiments, the image noise suppression apparatus 10 is configured in a terminal device, such as but not limited to a mobile phone, a tablet computer, a notebook computer, a desktop computer, a camera device or a smart wearable device.

In some embodiments, the image noise suppression device 10 further includes an image display device and a storage device. The image display device is used for displaying the noise suppression image IM 2. The image display device is, for example, but not limited to, a liquid crystal display, a light emitting diode display, an organic light emitting diode display. The storage device is used for storing the test image IM0, the original image IM1 and the noise suppression image IM2, and storing the above-mentioned various values and signals. Such as but not limited to volatile memory, read only memory, flash memory, magnetic disk.

In some embodiments, each image pixel PX1 can have not only one image pixel value, but also image pixel PX1 can have a variety of image pixel values, i.e., image pixel values such as, but not limited to, red light, green light, blue light, infrared light, etc. The image noise suppression apparatus 10 and the image noise suppression method can be adapted to different color spaces, such as but not limited to Gray scale (Gray) color space, RGB color space, YUV color space, RAW format color space, and the like.

In summary, in some embodiments of the present disclosure, the image noise suppression apparatus and the method thereof can obtain each corresponding noise suppression image pixel value according to the fixed noise value and the corresponding random number of each image pixel in the original image, and then obtain the noise suppression image according to the noise suppression image pixel value. The image noise suppression device and the method thereof obtain a shifted image pixel value, a shifted fixed noise value and a random number according to the bit number of the decimal of the fixed noise value, the image pixel value and the fixed noise value, obtain a shifted noise suppression image pixel value according to the shifted image pixel value, the shifted fixed noise value, the random number and a suppression equation, and obtain a noise suppression image pixel value according to the bit number of the decimal of the fixed noise value and the shifted noise suppression image pixel value. In some embodiments, an image noise suppression apparatus and method thereof perform a random function according to the number of bits of the decimal of a fixed noise value to obtain a random number, perform a decimal displacement on an image pixel value to obtain a shifted image pixel value, perform a decimal displacement on the fixed noise value to obtain a shifted fixed noise value, and perform a decimal displacement on the shifted noise-suppressed image pixel value to obtain a noise-suppressed image pixel value. In some embodiments, the image noise suppression device and the method thereof can utilize random numbers and decimal point displacement to avoid image distortion of an original image in the process of suppressing fixed pattern noise, especially image distortion generated by neglecting decimal point calculation of a fixed noise value. Therefore, the image noise suppression apparatus and the method thereof can obtain a noise-suppressed image with low fixed pattern noise.

Although the present disclosure has been described with reference to the preferred embodiments, it is not intended that all modifications and variations that fall within the spirit of the disclosure be included within the scope of the present disclosure.

Description of the reference numerals

Image noise suppression device

Image noise suppression device

100 processing circuit

300 image capturing circuit

SC scene

IM0 test image

IM1 original image

IM2 noise suppressed image

PX0 test pixel

PX1 image pixels

C0 line

D1 first axial direction

D2 second axial direction

S210-S250, S222-S226, S410-S440