TWI443605B - Method for adjusting color value of pixel in video signal - Google Patents

Description

Image signal pixel point color value adjustment method

The present invention relates to a method for adjusting a color value of a pixel of a video signal, in particular, a method for adjusting a color value of a pixel in an image between an image signal and an image to filter out possible noise and improve the image. Output the quality of the display.

Please refer to the first figure, which is a block diagram of an image processing program. The image processing program is a technical feature that is used in the processing of a video signal (video) in a current general computer system. The image processing program of the technology first inputs the image signal into the computer system (block). B11), that is, the computer system accepts the image signal to be output and output, and then the computer system or the display decodes the image signal (block B12), so that the relevant processing unit in the computer system The image decoding process can be performed on the image signal, and after the processing is completed, the subsequent display output process can be performed through a display (block B13).

However, in the above signal transmission process, the noise may be mixed into the video signal at any time, which may affect the quality of the image displayed at the last display. Therefore, how to improve the signal-to-noise ratio (S/N, that is, the ratio of signal (Signal) and noise (Noise) of the final output image, in dB), can provide better image output quality and display effect. This is the main purpose of the development of this case.

The object of the present invention is to provide a method for adjusting the color value of a pixel of a video signal, which is applied to the color value adjustment of pixels in an image between an image signal and an image, and is calculated for each pixel by a dynamic detector. The degree of motion, and then different noise filtering processes according to the different degrees of motion of the pixels, can improve the quality of the image output display.

The present invention relates to a method for adjusting a color value of a pixel of a video signal, which is applied between a first image and a second image in an image signal, the adjustment method comprising the steps of: utilizing the second in the image signal A second pixel in the image performs a motion state detection on a first pixel in the first image of the image signal to obtain a motion level value; the motion level value is less than a first threshold value And performing a first color value adjustment on the first pixel; performing a second color value adjustment on the first pixel when the motion degree value is greater than a second threshold value; and And a third color value adjustment is performed on the first pixel when the first threshold value is between the second threshold value; wherein the first threshold value is a static threshold value, and the second threshold value is a motion threshold value.

According to the above method, the first image is a to-be-processed image existing in the image signal.

According to the above method, the motion state detection system includes the following steps: when the motion degree value is less than the static threshold value, determining that the first pixel appears as a stationary state; when the motion degree value is greater than the motion threshold value, Determining that the first pixel is presented as a motion state; and determining that the first pixel is in a half motion state when the motion level value is between the static threshold value and the motion threshold value.

According to the above method, the stationary threshold value and the motion threshold value can be set, and the static threshold value is smaller than the motion threshold value.

According to the above method, the second image is the first image adjacent to the first image existing in the image signal.

According to the above method, the motion state detection further includes the following steps: when the motion degree value is less than a first experience threshold, the motion degree value is calculated as a first function motion degree value; when the motion degree value is greater than a second empirical threshold value, the motion degree value is calculated as a second function motion degree value; and when the motion degree value is between the first experience threshold value and the second experience threshold value, The exercise degree value is calculated as a third function exercise degree value; wherein the first experience threshold value and the second experience threshold value may be set, and the first experience threshold value is less than the second experience threshold value, And the first function motion degree value, the second function motion degree value, and the third function motion degree value are in a predetermined range.

According to the above method, the motion state detection further includes the following steps: low-pass filtering the motion degree value, the first function motion degree value, the second function motion degree value, or the third function motion degree value, and The high frequency portions of each of them can be filtered out.

According to the above method, the first color value adjustment system can filter out a still pixel noise.

According to the above method, the static pixel noise filtering system comprises the steps of: averaging the color values of the first pixel in the first image and the second pixel in the second image in the first image, And the second pixel of the corresponding position in the second image is presented as the motion state; and the second pixel and the corresponding position in the first pixel and the second image in the first image The third pixel of the corresponding position in the third image performs numerical averaging of the color values, and the second pixel of the corresponding position in the second image is presented as the quiescent state, and the corresponding position in the third image The third pixel system appears as the motion state.

According to the above method, the second image is a first image adjacent to the first image existing in the image signal, and the third image is present in the image signal. The second image adjacent to the first image.

According to the above method, the second color value adjustment system may filter a motion pixel noise, and the moving pixel noise filtering step includes: retaining the color of the first pixel in the first image value.

According to the above method, the third color value adjustment system can filter out half of the motion pixel noise.

According to the above method, the semi-moving pixel noise filtering system comprises the following steps: performing the still pixel noise filtering on the first pixel in the first image to obtain a first still pixel noise filtering. a value; calculating the motion level value and the static threshold value, the motion threshold value to obtain an interpolation coefficient; and performing, by using the color value of the first pixel, the first still pixel noise filtering value, and the interpolation coefficient The values are averaged to obtain a first half-motion pixel noise filtering value.

According to the above method, the image signal system comprises a plurality of images, and each of the images comprises a plurality of pixels, and the method is applicable to the pixels of the images.

The method used to eliminate noise in this case mainly uses the following two basic arithmetic processing methods, namely inter-field interpolation and intra-field interpolation. . Since the motion picture or the animation is composed of a series of images or image frames of the frame, the processing of the noise filtering is to perform the calculation of the internal pixels of one or more of the images. Then, the inter-image interpolation method can also be called temporal interpolation, because most of the general noises have high frequency characteristics and other surrounding low frequency pixel contents in the image. There are obvious differences, and the correlation of noise in the time direction between images is low, so the temporal interpolation method can use this feature to frame adjacent frames in the frame memory. Pixels at the same position of one or more images are averaged (usually for linear averaging), while high-frequency noise portions that are uncorrelated in the time direction are filtered to preserve the desired low frequencies. Information section.

In addition, the intra-image interpolation method can also be called spatial interpolation. Also, due to the high-frequency characteristics of the noise, it is obvious in some images and other surrounding low-frequency pixel contents. The difference is that the noise is less correlated in the spatial direction of the image. With this feature, the spatial interpolation method uses the noise and the surrounding low-frequency pixels in the image to be processed. The averaging process (usually for linear averaging) is performed, while the high-frequency noise portions that are uncorrelated in the spatial direction are filtered to preserve the desired low-frequency information portion.

As far as the temporal correlation of pixels between images is concerned, the temporal interpolation method will have better results in the noise processing on the static image, and the spatial interpolation in terms of the spatial correlation of the pixels in the image. The method of noise processing on the motion picture will have better results. The above two calculation methods only have adaptability in time domain and space domain, but still can not accurately display the authenticity of the image; therefore, in order to produce better calculation results, combine this A motion adaptive algorithm for both methods is proposed; this method uses a motion detection to detect the degree of motion of pixels within an image, but for motion or Non-motion conditions to properly select the results of the calculation using spatial interpolation or temporal interpolation; as shown in the second figure, it is a schematic diagram of the operation of the dynamic adaptive algorithm, which mainly uses a multiplex The timer 20 selects a temporal interpolation method or spatial interpolation of a temporal interpolation module 22 or a spatial interpolation module 23 according to the input condition of the image signal and the detection result of the motion detector 21. The result of the calculation of the method, and the high-frequency noise portion of the image signal can be filtered out and then output.

However, the above description is not only a temporal interpolation method using temporal correlation of pixels between images, but also a spatial interpolation method for spatial correlation of pixels in an image, or making full use of the existence time. Dynamic adaptive algorithm for the correlation between domain and spatial domain, where there are still calculus instability and defects in the final output display, because between the images of the sequence or between pixels within an image Originally, it will have a strong correlation, but if the information presented between images or between pixels varies greatly, it may not be able to distinguish the details of the image or the high-frequency noise; for example; : In the calculation of the temporal interpolation method, the time-averaged or high-frequency filtering of the pixels appearing as the motion state is also performed, and the calculus of the spatial interpolation method and the dynamic adaptive algorithm is performed. , the pixel content that appears to be strenuous motion change in the original image information may be ignored, which may cause the required image information to be erased together, so that the image is blurred and unclear. The case of loss or image picture detail, resulting in the final of the resulting overall picture becomes untrue. Therefore, the main purpose of the development of this case is to improve the problem of this conventional technology.

In the preferred embodiment of the present invention, we propose a method for adjusting the color value of the pixel of the image signal to solve the above-mentioned defect; the preferred embodiment can be applied to the related image processing chip and image processing program in the conventional computer system. An image signal is processed during image processing.

Please refer to the third figure, which is a block diagram of an image processing program proposed in this case. As can be seen from the figure, the image processing program first inputs the image signal into the above-mentioned conventional computer system (block B31), that is, the computer system accepts the image signal to be output and output, and then the computer system or display a program for decoding the image signal (block B32), so that the relevant processing unit in the computer system or display can perform desired image processing on the image signal, and therefore, a post-processing program (Post-Processing) (Block B33) can be used to perform the above image processing process, including image detection, de-interlacing, etc., and the color value of the pixel of the image signal proposed by the preferred embodiment of the present invention The adjustment method is also completed in this process. When the image processing is completed, the subsequent display output process can be performed (block B34).

Please refer to the fourth figure, which is a schematic diagram of a pixel signal color value adjustment process of the image signal using the method of the present invention. In the preferred embodiment of the present invention, the image signal, that is, the motion image or the animation as described above, is composed of a series of image images of a frame or a frame, that is, a plurality of images exist in the image signal, and Each of the images each contains a plurality of pixels.

The feature of the present invention is that we will first perform a motion state detection on each of the pixels to be processed in the image signal, such as an image F10 shown in the fourth figure, to determine the The value of the motion level of the pixel is determined by using the correlation of the aforementioned continuous image in the time direction, that is, one image to be processed (or one of the current images) and one In the previous image comparison, the previous image in this embodiment is the first image adjacent to the image F10 existing in the image signal, as shown in the figure, It is an image F11, and since the image F10 and the image F11 have the same frame range (or array size) definition, the pixels formed therein also have corresponding positions (or coordinates). Relationship, so as long as each pixel on the image F10 and each pixel at the corresponding pixel point position on the image F11 are compared between the two (pixel color value) differences, we It can be known that each pixel on the image F10 is in each of the image signals. The value of the degree of motion between images.

The motion state detection, that is, the calculation of the motion degree value. In this embodiment, a single pixel a in the image F10 is taken as an example, and the corresponding position on the image F11 of the previous image is used. (ie, the same coordinates as each other) is a pixel a1, we subtract the color value of the pixel a and the pixel a1 and then take the absolute value, so that the difference between the two pixels can be represented by a positive value, and the value is The size also represents the degree of motion (defined as m) exhibited by the pixel on a certain coordinate when the image signal is from the image F11 to the image F10. The present invention is characterized in that, in order to clearly divide the generated motion degree value, we further introduce a saturation function, which uses the two empirical threshold values α, β for the generated motion degree value m. It is divided into three kinds of function operations, and its function operation results are as follows:

Here, m is the original motion degree value, and m' represents the motion degree value after the function operation, wherein when the original motion degree value m is less than or equal to the experience threshold value α, the motion degree value is calculated as 0, when the exercise degree value m is greater than the experience threshold value β, the exercise degree value is calculated as 255, and when the exercise degree value is greater than the experience threshold value α and less than or equal to the experience threshold value β, The motion degree value is calculated as ((m-α)/(β-α))*255; the two empirical threshold values α, β can be set here, and the empirical threshold α is less than the empirical threshold β, In this embodiment, we can set the empirical threshold α to be 5, or between 5 and 10, and the empirical threshold β can be set to 100, or between 60 and 150, and this When the color value of the 8-bit (bit) is displayed, the last obtained motion level value m' can be set between 0 and 255.

In this way, through the calculation and application of this saturation function, we can filter out the influence of high-frequency noise on the motion state detection, and normalize the obtained result in three functional operation modes. With the use of the degree function, we can determine the magnitude of the motion level value in a predetermined range, and if the operation of the saturation function is not used, the magnitude of the value obtained after the motion state detection cannot be determined. In what range (ie, the maximum and minimum values cannot be determined).

Furthermore, we can further low-pass filter the obtained motion degree value m' (or the original motion degree value m), and the low-pass filter can be a matrix operator, and After low-pass filtering, a more accurate motion level value m" can be obtained, which can be: The obtained motion degree value m" is obtained by weighting the motion degree value m' (or the original motion degree value m) of the pixel to be processed and the motion degree values of other surrounding pixels, thereby being able to increase the height thereof. The frequency portion is filtered out so that a more accurate pixel motion level value can be obtained.

As described above, we use this motion state detection to detect each pixel in an image. As in the fourth figure, we can know from the image F11 to another image F10. The value of the degree of motion of each of the pixels; as described above, or according to the current technique for displaying the output of the image, there is a certain temporal correlation or spatial correlation between consecutive images or within an image. Therefore, we can judge whether the pixel belongs to a stationary state or a motion state in the time domain between images according to the obtained pixel motion degree value.

The feature of the present invention is that when we judge that the pixel is in a stationary or moving state, half of the motion state is added. As the name implies, the semi-moving state is a state between the stationary state and the motion state, and the judgment is made. The state of the pixel is the motion degree value obtained by detecting the motion state (which may be the original motion degree value m calculated by the unsaturation function, or the motion degree value m' calculated by the saturation function. Or, the motion degree value m", which is calculated by the low-pass filter, is compared with a motion threshold value S and a motion threshold value M, when the motion degree value m is compared. 'When it is less than or equal to the static threshold value S, it is judged that the pixel appears as the stationary state, and when the motion degree value m' is greater than the motion threshold value M, it is determined that the pixel appears as the motion state, and when the motion degree value is When m' is greater than the static threshold value S and less than or equal to the motion threshold value M, it is determined that the pixel appears to be in the semi-motion state.

Here, the static threshold value S and the motion threshold value M can be set, and the static threshold value S is smaller than the motion threshold value M. In this embodiment, we can set the static threshold value S to 30. , or between 10 and 40, and the motion threshold M can be set to 80, or between 60 and 150; therefore, for different degrees of motion, we can be on the image to be processed All pixels are judged to belong to a stationary state, a motion state, or a semi-motion state in the time domain between the images.

As described above, in this embodiment, after the motion state detection is performed, and the motion degree values of the pixels are classified, the present invention is characterized in that we adopt different color value adjustments for pixels of different states. In order to display the best image display output in motion pictures or animations. Please continue to refer to the fourth figure. There are pixels a, b, and c in the image F10, and the pixels a, b, and c are the pixels to be processed, and the corresponding positions in the image F11 (or There are pixels a1, b1, and c1 on the coordinates. We assume that in this embodiment, the pixels a, b, and c represent the stationary state, the motion state, and the half motion state respectively after the motion state detection, and respectively Performing a still pixel noise filtering on the pixel a for the stationary state, performing a motion pixel noise filtering on the pixel b for the motion state, and performing half-motion pixel noise on the pixel c in the semi-motion state. The color value adjustment of different ways, such as filtering, is adjusted as follows.

In this embodiment, the pixel a is in a stationary state, and the pixel a in the image F10 in the fourth figure and the pixel a1 in the image F11 are pixels corresponding to the position, and the present case The process of static pixel noise filtering is to sequentially process the pixels of the previous one or more images into the color values according to the display order of the images in the static state, and the main concept is that All pixels of the corresponding position in the state before the pixel a are color-averaged numerically averaged, and averaged to pixels appearing as motion states at corresponding positions in the previous image.

For example, the pixel a (the color value of which is also denoted by a) in the image F10 of this embodiment is presented as a stationary state, and the pixel a1 (the color value thereof is also represented by a1) in the image F11 is presented as a motion. The state, if it is assumed that the value of the pixel a after filtering through the still pixel noise is a', the average value of the color value of a' is calculated and the result is (a+a1)/2. Similarly, if the pixel a1 in the image F11 is also in a stationary state, but an image in front of the image F11 (ie, the second image adjacent to the image F10) When the pixel a2 of the same corresponding position in the image (not shown in the figure, the color value is also represented by a2) is in the motion state, it is assumed that the value of the pixel a after the static pixel noise filtering is also a', Then the average value of the color value of a' is calculated and the result is (a+a1+a2)/3. However, if the pixel a2 is also in a static state, then more previous images need to be included to perform the pixel color value. On average, in this embodiment, we count by 8-bit (bit) number of vectors, so we can make a numerical average of the total number of pixels at a maximum of 256 times, or you can use other comparisons. The small bit or the larger bit is used to count the number of times, and the average of the static states of the total number or the total number can be used separately.

Therefore, according to the calculation of the numerical average of the color values, it is possible to use the pixel information in many images before the image to be processed, however, the related image processing in the conventional computer system or display The space of the memory is limited, and it is impossible to record all the processed images. Therefore, for this situation, we introduce a calculation method of the vector of the number of still periods of the pixel, that is, the numerical value averaging of the color values by the above method. After the calculation, the number of occurrences of the quiescent state accumulated by each pixel is recorded. For example, when the pixel a is in a stationary state and the pixel a1 is in a motion state, the number of quiescent states is defined as one time, if the pixel a, When a1 is in a stationary state and the pixel a2 is in a motion state, the number of stationary states is defined as 2 times, and so on. Therefore, as long as the necessary pixel information of the previous image of the image to be processed is recorded, and the color value is averaged for the image to be processed, the corresponding static state number can be used for calculation. This allows the above average method to be completed with only the necessary information recorded.

For example, when the pixel a, a1 is in a stationary state and the pixel a2 is in a motion state, the number of quiescent states is 2, and the value a' of the pixel a after the static pixel noise filtering is ( a+a1+a2)/3, however, if it is assumed that the value of the pixel a1 after the static pixel noise filtering is a1', then the average value of the color value of a1' is calculated and the result is (a1+a2)/2, therefore, a' can be represented by (a+2*a1')/(2+1), and 2 of them are the number of quiescent states, so only the necessary pixel information of the previous image of the image to be processed (ie, noise filtering) In addition to the value of the latter) can be recorded.

Next, in this embodiment, the pixel b is in a motion state, and the pixel b in the image F10 in the fourth figure and the pixel b1 in the other image F11 are pixels corresponding to the position, That is, the pixel b exhibits a motion state with respect to the pixel b1; since the pixels appearing in the motion state are in a sequence before and after the output of the image is displayed, there is a certain degree of change in the color value, and such a change can be made for the human eye. The degree of observation is indistinguishable, that is to say, such pixel changes are less sensitive to the observation of the human eye. Therefore, the feature of the present case is that we perform the motion on the pixel b presented as a motion state. Pixel noise filtering, the main concept is to retain the color value of the pixel b in the image F10, that is, not to perform any actual operation or numerical average adjustment on the pixel b, in order to retain the detailed features of the color value change therein. Therefore, it is possible to avoid a phenomenon in which image blurring or missing details may occur when pixels between images or images are presented in a moving state.

Furthermore, in this embodiment, the pixel c is in a semi-motion state, and the pixel c in the image F10 in the fourth figure and the pixel c1 in the image F11 are pixels corresponding to the position, That is, the pixel c exhibits a semi-motion state between a stationary state and a motion state with respect to the pixel c1. The process of filtering the half-motion pixel noise in the feature of the present invention firstly performs the static pixel noise filtering as described above for the pixel c in the image F10 (in the same manner, that is, forward average to A pixel of the motion state is not described here, and a still pixel noise filtering value c' is obtained; then, the motion value m of the pixel c (whether m, m' or m) is A variable having a different result from different degrees of calculation or classification of different pixels, where the calculation is performed for the pixel c and the pixel c1, and the static threshold value S and the motion threshold value M are performed. Calculating to obtain an interpolation coefficient f, in this embodiment, the calculation and result of the interpolation coefficient f can be (m-S) / (M-S); finally, the color value of the pixel c is utilized (at the same time) The value of the pixel c is filtered by the motion pixel noise, and the color value is also represented by c. The static pixel noise filtering value c′ and the interpolation coefficient f are numerically averaged, and finally Half of the motion pixel noise filtering value, and in this embodiment, the half motion pixel is mixed The calculation and result of the filtering value can be: c*f+c’*(1-f).

Therefore, for each pixel in an image to be processed, we can perform the motion state detection and the corresponding pixel point color value adjustment as described above, including three levels of static, motion, and semi-motion. The state of the pixel noise filtering is performed by using the calculation method adopted in the preferred embodiment of the present invention, so that an appropriate pixel color value adjustment result can be obtained. Further, in this embodiment, when noises that may exist in the image of the image to be processed are filtered out, it is necessary to perform corresponding adjustment on each pixel in the image, and finally obtain the obtained The image to be used for display output is the static pixel noise filter value and the motion pixel noise filter value of each pixel obtained after the adjustment (in this case, the color value of the pixel to be processed) And/or the result of the half-motion pixel noise filtering value is output and displayed.

The flow of the preferred embodiment of the present invention can be summarized as shown in the flowchart of FIG. First, the image to be processed (ie, the current image) is received (step S10), and the previous image of the image to be processed is temporarily stored (step S11), and then the two images are performed. a motion state detection (step S12), thereby being able to obtain a motion degree value corresponding to each pixel in the image to be processed (step S13), and further determining that the pixel system is in a stationary state according to the motion degree value, a motion state or a half motion state (step S14); secondly, when the pixel is in a stationary state, performing a still pixel noise filtering operation process (step S15), when the pixel is in a motion state, An operation process of moving pixel noise filtering (step S16), and when the pixel is in a semi-motion state, performing half-motion pixel noise filtering operation processing (step S17); finally, according to step S15 to step As a result of S17, the image filtered by the noise is output and displayed (step S18).

However, the relevant threshold values used in the present embodiment, for example, the empirical threshold values α, β, the static threshold value S, the motion threshold value M, etc., can all be adjusted within a certain range, and the pixels of the obtained image are obtained. The point color value adjustment can also be similar to the embodiment; at the same time, the saturation function and the low-pass filtering process used in the present embodiment can also be omitted, and the motion state detection can be directly used. The resulting motion degree value is used for subsequent judgment and calculation, or the operation of the saturation function and the low-pass filtered matrix operator may be replaced by other numerical values or calculation methods, for example, the low-pass filtered matrix. The operator is replaced by: Or the processing of the low-pass filter matrix operator from low-order (for example, 3×3) to high-order (for example, 5×5) and the like are all changes that can be associated under the concept of the method. Furthermore, the half-motion pixel noise filtering used in the embodiment may be changed, wherein the calculation method of the interpolation coefficient f may also be changed, for example, the obtained motion degree value m is directly divided by 255. That is, f=m/255, or calculate the brightness value of the pixel itself to obtain the interpolation coefficient, that is, f=source(x, y)/255, where source(x, y) is the pixel Brightness value.

In summary, the method of the present invention can effectively filter out high-frequency noise in the image, and at the same time, the processing application of the semi-moving pixel of the present invention is added, so that the adjustment of the color value of the pixel can produce a smooth transition. Therefore, it is possible to avoid the problem that the noise generated in the prior art is mixed into the image signal, and solve the problem of image distortion caused by the image blurring, unclearness or missing details caused by the conventional adjustment technique, so that the development of the case is successfully achieved. The main purpose.

The present invention has been modified by those skilled in the art and is intended to be modified as described in the appended claims.

The components included in the diagram of this case are listed as follows:

Multiplexer. . . 20

Motion detector. . . twenty one

Temporal interpolation module. . . twenty two

Spatial interpolation module. . . twenty three

Square. . . B11~B13, B31~B34

image. . . F10, F11

Pixel. . . a, b, c, a1, b1, c1

step. . . S10~S18

This case can be obtained through a more detailed understanding of the following figures and descriptions: The first figure is a block diagram of an image processing program.

The second picture is a schematic diagram of the operation of the dynamic adaptive algorithm.

The third figure is a block diagram of an image processing program.

The fourth figure is a schematic diagram of the process of adjusting the color value of the pixel of the image signal by the method of the present invention.

The fifth figure is a flow chart of a preferred embodiment of the present invention.

step. . . S10~S18

Claims (14)

A method for adjusting a color value of a pixel of a video signal is applied between a first image and a second image in an image signal, the method comprising the steps of: utilizing the second image in the image signal a second pixel to perform a motion state detection on a first pixel in the first image of the image signal to obtain a motion level value; when the motion level value is less than a first threshold value, The first pixel performs a first color value adjustment; when the motion level value is greater than a second threshold value, performing a second color value adjustment on the first pixel; and the motion level value is between the first threshold And between the value and the second threshold, performing a third color value adjustment on the first pixel; wherein the first threshold is a static threshold and the second threshold is a motion threshold. The image signal pixel color value adjustment method of claim 1, wherein the first image is a to-be-processed image existing in the image signal. The image signal pixel color value adjustment method of claim 1, further comprising the steps of: determining that the first pixel appears to be a stationary state when the motion degree value is less than the static threshold value; When the motion degree value is greater than the motion threshold value, determining that the first pixel is presented as a motion state; and when the motion degree value is between the static threshold value and the motion threshold value, determining that the first pixel is presented as a half Movement state. The image signal pixel color value adjustment method of claim 1, wherein the static threshold value and the motion threshold value are set, and the static threshold value is smaller than the motion threshold value. The image signal pixel color value adjustment method according to claim 1, wherein the second image is a first image adjacent to the first image existing in the image signal. The image signal pixel color value adjustment method according to claim 1, wherein the motion state detection further comprises the following steps: when the motion degree value is less than a first experience threshold, the motion degree value is calculated. a first function motion degree value; when the motion degree value is greater than a second experience threshold value, the motion degree value is calculated as a second function motion degree value; and when the motion level value is between the first experience level When the threshold value is between the threshold value and the second empirical threshold value, the motion degree value is calculated as a third function motion degree value; wherein the first experience threshold value and the second experience threshold value may be set, and the The first empirical threshold value is less than the second empirical threshold value, and the first function motion degree value, the second function motion degree value, and the third function motion degree value are in a predetermined range. The image signal pixel color value adjustment method of claim 6, wherein the motion state detection further comprises the following steps: the motion degree value, the first function motion degree value, and the second function motion degree. The value or the third function motion level value is low pass filtered, and the respective high frequency portions can be filtered out. The image signal pixel color value adjustment method of claim 3, wherein the first color value adjustment system is a static pixel noise filtering. The image signal pixel color value adjustment method of claim 8, wherein the still pixel noise filtering system comprises the following steps: the first pixel and the second image in the first image. The second pixel of the corresponding position performs a numerical average of the color values, and the second pixel of the corresponding position in the second image is presented as the motion state; and the first pixel and the first pixel in the first image The second pixel of the corresponding position in the second image and the third pixel of the corresponding position in the third image perform numerical average of the color values, and the second pixel of the corresponding position in the second image is presented. For the quiescent state, the third pixel of the corresponding position in the third image appears as the motion state. The image signal pixel color value adjustment method according to claim 9, wherein the second image is the first image adjacent to the first image existing in the image signal, And the third image is a second image adjacent to the first image existing in the image signal. The image signal pixel color value adjustment method of claim 1, wherein the second color value adjustment system can filter a motion pixel noise, and the motion pixel noise filtering step includes : retaining the color value of the first pixel in the first image. The image signal pixel color value adjustment method according to claim 9, wherein the third color value adjustment system can filter out half of the motion pixel noise. The image signal pixel color value adjustment method according to claim 12, wherein the half-motion pixel noise filtering system comprises the following steps: performing the static pixel impurity on the first pixel in the first image. Filtering to obtain a first still pixel noise filtering value; calculating the motion level value and the static threshold value, the motion threshold value to obtain an interpolation coefficient; and using the color value of the first pixel, The first still pixel noise filtering value and the interpolation coefficient are numerically averaged to obtain a first half-motion pixel noise filtering value. The image signal pixel color value adjustment method of claim 1, wherein the image signal includes a plurality of images, and each of the image systems each includes a plurality of pixels, and the method is Applied to the pixels of the images.