TWI357583B - Method of correcting bad spots for digital camera - Google Patents

Description

IX. Description of the Invention: [Technical Field] The present invention relates to a method for solving a dead pixel on a digital image and a digital camera device having a dead pixel automatic correction function, which is applied to a digital camera device, The digital camera device takes a series of sample papers each having a different single grayscale value to generate a digital image of the sample paper, searching for a bad point coordinate and a dead pixel channel, and then using a pixel corresponding to the bad point coordinate. The intensity value of the color channel 'replaces the intensity value of the dead point. [Prior Art] Press, general digital camera equipment (such as: digital camera or mobile phone with digital camera or personal digital assistant), etc., including optical components (such as lenses, lens groups, etc.) and electronic components (such as: light sensitive) Detector, microprocessor and circuit components, etc.) 'When a user takes a scene with a digital camera, the light of the scene passes through the lens of the digital camera, focusing on the internal light sensor (eg, charge-collecting) On the component, referred to as Ccd, the optical sensor is used to convert the optical image of the scene into an electronic signal, and after processing, the digital image of the scene is obtained, and the digital image is made up of pixels (Pixels). Array is formed, and the number of pixels is related to the resolution of the digital camera device. Taking the general threat color mode as an example, each pixel has three color channels such as a red channel, a green channel, and a blue channel, and each color channel All have an intensity value, meaning that each digital image is a mixture of red, green and blue grayscale images (Gray Scale Image)Because digital images can be easily copied, transmitted or shared via the Internet, and saved through digital data, 'will not be yellowed due to age, with the popularity of home computers, digital storage methods Diversification, the maturity of electronic technology and the increase in market demand, digital camera equipment has become increasingly popular, among which, digital S-machine even replaces the traditional bottom; ί-style butterfly has become the mainstream of reducing the market, each _ manufacturing Wei f is not committed to improving products Age f and lower production costs, in order to gain a competitive advantage in the market. Since the manufacturing process of the optical components and electronic components of the digital camera device cannot guarantee the small alkali, the digital image of the digital image captured by the tilting device will have some pixels depending on the color and brightness and its target. The value is different. The so-called bad points of the image cut, usually the difference between the color and the brightness of the neighboring pixels is very large. Because the * influences the ηπ quality of the digital image captured, the digital camera equipment will do the bad points before leaving the factory. Check that when the number of dead pixels of a digital camera exceeds a certain allowable value, it is judged to be a defective product, and the defective product may need to be repaired and replaced with corrective measures, or directly scrapped, thus, resulting in manufacturing costs. Rising, even if the number of bad points does not exceed the allowable value, it will still be a complaint of a bad rider's complaint 'affecting the goodwill of the manufacturer'. How can we develop a search and correction digital position in today's highly competitive market environment? The method of the dead point on the photographic equipment, and the method is not only fast in the calculation process, but also can be easily written into a software body, and the various aspects are related to each other. Effective solution to the problem of missing dots on the digital image produced by each of the digital camera equipment, namely digital set __'s difficult committed to research and solve one important Tibetan issue. [Summary of the Invention] There is an increase in the cost of the occurrence of a bad point on the silk health (such as ··············································································· Research and experiment, the final Wei Xianben hair _ correction method on the number of bad points on the health equipment and the automatic correction of the number of miscellaneous equipment with the axis, which is different from the majority of the age of the county, to determine the dead point 'Without the total miscellaneous traits of the remaining mine image, this book uses the statistical method to get the total miscellaneous traits of the digital image, and then find all the none of the photographic equipment according to the overall feature scale, in order to Effectively search for the number of health equipment on the 'Wei timely correction, so that these bad points do not affect the digital image captured too much, even to let the user feel the effect of the new equipment. The present invention provides a method for correcting a dead pixel on a digital camera device, the method being a digital assistant device comprising photographing a sheet having a single gray scale value using the digital camera device 'To create a new image of the paper. - Generally speaking, through the digital image of the good digital... and the sighing of the paper, the color and brightness of each pixel and its neighboring pixels will not be much different, and the number of dead pixels will be different. The digital image generated by the camera _ is 'very different from the color and brightness of the neighboring pixels at the bad point. The cause of these dead pixels is the red channel and the green channel on the dead point. The blue channel is broken. It is also possible that a certain channel (such as a green channel) is broken, so the damaged color channel must be found. In addition, in theory, a single gray scale value of the sample 1575583 paper digits & image of each pixel of the three color channels have the same intensity =, but due to the variation of the silk component and the electronic component itself, and the ring i see The intensity of the three color channels of each pixel of the digital image is not the same, but a normal distribution, and the average value and the difference of the normal distribution are calculated by the method of statistical analysis. Under normal circumstances, the distribution of the intensity values of each color channel of all pixels should be within the range of the average & (such as: 8 times the standard deviation), in the case of a bad point, - a bad point at least The intensity value of the color channel is outside the range of the standard. Therefore, by using the sheet-like gray-scale sample paper, it is possible to find the - part of the dead pixels, with the series being different. With a single grayscale sample, you can find all the dead pixels. The method of the present invention comprises counting the intensity values of the color channels of all the pixels on the digital image, and calculating the average value and the standard deviation of the intensity values of the color channels, and multiplying by the standard deviation of the intensity values of the color channels. A multiple (eg, 8 times) 'as the threshold of the color channel (Thresh〇ld Umit Value, TLV for short), 嗣, calculate the deviation of the intensity values of the color channels of all the pixels on the digital image, and then judge Whether the difference between the intensity values of the color channels of all the pixels is greater than the threshold of each color channel, and if so, the pixel is a dead pixel, and the color channel is a bad point channel, that is, the bad point coordinate and the bad point channel are output. Thus, using the digital camera device to take a series of sample papers having different single grayscale values, to generate a digital image of the sample papers, "repeating the foregoing steps", respectively performing statistics and calculations on each of the digital images, ie For all pixels, search for all bad point coordinates and dead point channels' and all the dead pixels and dead pixels, record 9 for the digital camera Long features to serve as a subsequent correction of the digital image based on digital photographic apparatus Xu captured. Another object of the present invention is to determine whether there is a dead pixel channel in each color channel of each pixel according to the feature data of the digital camera device, and if so, the intensity value of a bad dot channel is used for the dead pixel The intensity value of the corresponding color channel of a surrounding pixel is replaced, or is replaced by one of the intensity values of the corresponding color channels of the plurality of pixels around the dead point, and the interpolation is among the intensity values of the corresponding color channels. A value between the maximum and minimum intensity values, repeating the foregoing steps until all pixels on the digital image complete the determination of the dead pixels and the correction process. The method for searching and correcting dead pixels on a digital camera device is not only simple and fast, but can be widely applied to various digital camera devices by being programmed into software to instantly correct the digital position generated by each digital camera device. A dead pixel that occurs on the image. It is still another object of the present invention to provide a digital photographic apparatus having a dead-end automatic correction function, the digital photographic apparatus having a dead-point processing software capable of relying on stored dead-point coordinates and bad-point trajectories, Determining whether the color channel of each pixel is a bad point channel on a digital image generated by the digital camera after shooting a scene, and using the intensity value of a bad point channel with the corresponding number of pixels around the dead point The interpolation of the intensity value of the color channel is replaced, so that the digital camera can correct the dead pixel in real time. In order to facilitate the review committee, the principles and functions of the method of the present invention can be further understood and understood. The embodiments are illustrated and described in detail with reference to the following drawings: [Embodiment] The present invention provides a digital camera for correcting digital images. A method of applying a dead pixel to a digital camera device. In a preferred embodiment of the method, 'please refer to FIG. 1 to take a single grayscale value with the digital camera device 2 The sample paper 1〇〇 (eg, all white paper) is used to generate a digital image 200 of the sample i〇i, each pixel of the digital image 200 has a red channel, a green channel and a blue channel. Wait for three color channels, each color channel has an intensity value. Under normal circumstances, the color of the digital image 200 is displayed in the same manner as the sample 100, and if there is a bad point, a variation occurs at a bad point, such as a dark spot. In this embodiment, the intensity values of the color channels of all the pixels on the digital image 2 are counted to calculate the average value and standard deviation of the intensity values of the color channels, as shown in FIG. 2 The statistics and calculation steps are described in detail as follows: (601) The sum of the intensity values of the red channel is represented by Sum-R, the sum of the intensity values of the green channel is represented by Sum_G, and the sum of the intensity values of the blue channel is represented by Suin_B 'red channel The sum of the squares of the intensity values is represented by Sum_R2. The sum of the squares of the intensity values of the green channels is represented by Sum_G2. The sum of the squares of the intensity values of the blue channels is represented by Sum_B2, and the initials of Sum_R, Sum_G, SumJB, Sum_R2, Sum_G2, and Sum_B2 are set. Value (eg, initial value is 0); (602) Enter the intensity of the three color channels of a pixel of a digital image by 1575583 (603) Let Sum_R = Sum - R + the intensity value of the red channel, so that Sum_G = Sum_G + green channel Intensity value, let Sum__B=Sum_3+ the intensity value of the blue channel, let Sum-R2=Sum_R2+ (the intensity value of the red channel) 2, let Sum-G2=Sum-G2+ (the intensity value of the green channel; Γ2' to Su m—B2=Sum_B2+ (the intensity value of the blue channel; r2; (604) Determine whether the intensity values of the three color channels of all pixels have been input 'if yes' into the step (6〇5), if not, go to the step ( 6〇2), to input the intensity values of the three color channels of the next pixel;

(605) Calculate the average value of the intensity values of the red, green, and blue channels AV-R, AV-G, AV-B', and the standard deviation STD_R of the intensity values of the red, green, and blue channels, STD_G, STD_B, and calculate the threshold of the red channel, the green channel and the blue channel, and set the total number of pixels to pixs, then: AV_R = Sum_R/pixS > AV__G = Sum_G/pixs > AV - B = SumJB/pixs,

STD G

STD B STD R Sum _ _R2- 2 Sum R pixs pixs -1 Sum _G1 — Sum _ G2 pixs pixs -1 Sum _B2~ Sum B2 ----- pixs pixs -1 The red channel is close to the alpha of the STD_K. Alpha can be 12 1357583 through experiments to obtain 'the green channel threshold is the standard deviation STD-G alpha times' blue channel threshold is the standard deviation STD-B alpha times. In the embodiment +, 4, as shown in the third item, the steps of searching for the bad point coordinates and the dead point channel of the digital scene image 200 are described in detail as follows: (701) Input one pixel of red, green, blue, etc. The intensity values of the three color channels; (702) Calculate the difference between the intensity values of the red, green, and blue channels DR, DG, DB: DR = the intensity of the red channel - war _ 丨, ® =I intensity value of green channel -AV_G| , DB= | intensity value of blue channel - Αν_β| ; σ〇3) Determine whether the deviation of the intensity value of the red channel exceeds the threshold of the red channel, and if so, The pixel is a dead pixel, the red channel is a bad point channel, and the process proceeds to step (704). If not, the process proceeds to step (705); - (7〇4) outputs the bad point coordinate and the bad point channel, and proceeds to the step (7〇). 5). (7〇5) Determine whether the intensity difference DG of the green channel exceeds the threshold of the green channel. If yes, the pixel is a dead pixel, and the green channel is a bad point channel, and the process proceeds to step 706. ), if not, proceed to step (707); (706) turn out the bad point coordinates and the bad point channel, enter the step (7ο?). (7) Determine whether the intensity difference DB of the blue channel exceeds the threshold of the blue channel. If yes, the pixel is a dead pixel, and the blue channel is a bad channel. Go to step (708), if no, go to step (709); 13 (708) output bad point coordinates and dead point channel, enter step (7〇9); and (709) determine whether all pixels have three color channels The intensity values are all rounded, if yes, if not, go to step (7〇1) to input the intensity values of the three color channels of the next pixel. In this embodiment, before the digital camera 2 is shipped from the factory, First, use the digital camera device 2 to take a series of samples from white to black, multiple sheets of different gray scale values to generate digital images of the samples, and use a computer device to perform the second In the steps of FIG. 3 and FIG. 3, statistics and calculations are performed on each of the digital images, and all the bad point coordinates and the bad point channels on the digital camera device 2 are searched for, and all the bad point coordinates and the bad point channel records are recorded. Characteristic data of the digital camera device 2, and the feature data It is stored in the digital camera device 2 as a basis for correcting the captured digital image by the subsequent digital camera device 2. In this embodiment, after the digital camera device 2 is shipped, the user takes a scene with the digital camera device 2 Thereafter, the digital camera device 2 determines and corrects the bad point coordinates and the dead pixel channel on the digital image of the scene according to the following steps: (810) inputting a pixel coordinate; (820) according to the characteristic data Determining whether there are dead pixels in the three color channels of the pixel. If yes, proceed to step (83〇), if not, proceed to step (840); (830) when any color channel of the pixel is a dead pixel In the channel, the intensity value of the corresponding color channel of one pixel of the plurality of pixels around the dead pixel is used to replace the intensity value of the bad point channel, or one of the intensity values of the corresponding color channel of the 1357583 plurality of pixels around the dead point is used. Interpolation replaces the intensity value of the dead pixel channel, so that the intensity value of the bad point channel can be corrected. 'The interpolation is the maximum and minimum among the intensity values of the corresponding color channels. a value between the intensity values, for example: the interpolation is an average of the intensity values of the corresponding color channels of the pixels surrounding the dead point; and (840) determining whether all pixel coordinates have been input, and if so, ending, if not Go to step (810) to enter the next pixel coordinate. The steps (10)) to (10) above are repeated until all the pixels in the _digit county complete the judgment of the dead point and the correction processing. In another embodiment of the present invention, a digital camera device 5 having a dead pixel automatic correction function is shown in FIG. 5. The digital camera device 5 includes a photosensitive member 51, an optical device 50, and an arithmetic processing unit 52. a bad point processing software 53, a memory 54 and a display screen 55, wherein the photosensitive element 51 is configured to convert an optical image of a scene into a digital image to obtain a three-color channel of the digital image towel. The optical device 50 is configured to project and focus an optical image of a scene on the photosensitive element 51. The arithmetic processing unit 52 performs data calculation and data processing according to the instruction of the dead pixel processing software 53, and performs data processing and data reduction. 53 money counts the intensity value of each color channel of each pixel on a digital image, calculates the intensity value of each color channel and the standard deviation of the intensity value of the color channel multiplied by a multiple, as the color The threshold of the channel, and the difference between the three red channel (four) degrees of each pixel is greater than the threshold of each color channel, and the intensity value of any color channel of any pixel is found 15 1357 If the averaging difference of 583 is greater than the threshold of each color channel, the pixel is a dead pixel. The coordinates of the pixel are marked as a bad point coordinate, and the color channel is marked as a bad point channel, and the threshold value is bad. Point coordinates and dead point channels are stored in the memory 54. Referring to Fig. 6, the user takes a series of photo papers 5 from a full white to a full black, and a plurality of sample sheets p each having a different single gray scale value to produce a digital image I of the sample paper P.嗣, the digital camera device 5 separately calculates the intensity values of the three color channels of each pixel of each digital image I, and performs the statistics and calculations in the foregoing method steps, thereby searching for all the bad point coordinates and dead pixels on the digital camera device 5. a channel for storing all of the dead pixel coordinates and the dead pixel channel of the digital camera device 5 in the memory 54 of the digital camera device 5 as a basis for the rear digital camera device 5 to continuously correct the captured digital image, and then The dead pixel processing software 53 can determine whether the digital camera device 5 is photographed according to the dead pixel coordinates and the dead pixel channel stored in the memory 54, and whether the color channels of each pixel are on a digital image generated after the object is Is a dead pixel channel, which is determined as the intensity value of a bad point channel, and is replaced by one of the intensity values of the corresponding color channels of the plurality of pixels around the dead point, wherein the interpolation is for each corresponding color channel Among the intensity values, a value between the maximum and minimum intensity values, so that the digital camera device 5 can automatically correct the dead pixels in real time, so that the display screen 55 can present the The digital image generated after the correction processing by the dead pixel processing software 53 is performed. According to the above description, the method of the invention not only is simple and fast, but can be easily edited into a soft body, which is widely applied to various digital camera devices to generate an effective and timely red color digital camera. The number of defects in the 1^7583 shirt image can be used not only for the inspection and correction of the digital camera equipment before the wei wei, but also when the new vital phase equipment has been used for a period of time. It can also be repaired by this method, and there is no need to replace the parts, that is, the damage point of the camera can effectively reduce the manufacturing and maintenance costs, and the user can take pictures after taking pictures of the owing-digital camera. The device can immediately correct the dead pixels of the digital image, so that the output digital image has no dead pixels, which greatly improves the convenience of use. The above is only the preferred embodiment of the present invention, but the scope of the claims of the present invention is not limited thereto, and those skilled in the art can easily think according to the technical content disclosed by the present invention. And equivalent changes, should be without departing from the protection of the invention. BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a schematic diagram of a digital image of a sample of a single gray scale value; FIG. 2 is a flow chart of the average value and standard deviation of the intensity values of each color channel of the present invention; 3 is a flow chart of the dead pixels of the digital photography device of the present invention; FIG. 4 is a flow chart of the dead pixels of the digital camera device of the present invention; FIG. 5 is a digital display of the automatic dot correction function of the present invention. The camera system diagram; and Fig. 6 is a schematic diagram of a series of digital images of sample papers having different single gray scale values. [Description of main component symbols] 17 1357583 Optical device............50 Photosensitive element............51 Operation processing unit......... ...52 Bad Point Processing Software............53 Memory.........54 Display Screen............ 55

Sample paper ............100, P digital camera equipment............2

Digital image ............200,I digital camera equipment............5

Is

Claims (1)

1357583 X. The scope of application for patents: 1. A method for correcting the materials of Shangga County, which is applied to a digital camera, including: using the digital camera to take at least a sample with a single-value value to obtain The intensity value of the three color channels of each pixel in the sample-digital image 'the two color channels are the red channel' green channel and the blue channel; and the intensity values of the two color channels of each pixel are calculated The average value and the standard deviation of the intensity values of the color channels' and calculate the threshold of the two color channels of the image, wherein the threshold of each color channel is equal to the standard deviation of the intensity values of the color channel multiplied by a multiple; Each of the pixels is used to determine whether each pixel has a dead pixel, and if so, the bad point is corrected by the color channel intensity value of the plurality of pixels around the dead point. 2. The method of claim 1, comprising calculating an average deviation of intensity values of each of the color channels of the pixel on the digital image; and determining an intensity of each of the color channels of the parent pixel. Whether the difference is greater than the threshold of each color channel. When the difference between the intensity values of any color channel of one pixel is found to be greater than the threshold of the color channel, the pixel is a dead pixel, and the coordinates of the pixel are marked as a bad Point coordinates, the color channel is marked as a bad point channel. 3. The method of claim 2, comprising: using the digital camera device to take a series of sample papers each having a different single grayscale value to generate a digital image of the sample paper and then separately digitizing the digital image , for statistics and calculations, 19 to search for the bad point coordinates and the bad point channel of the digital image, and then search for all the bad point coordinates and the bad point channel on the digital camera device. 4. If the method of pleading for the job 3 includes changing the coordinates of the bad point and the bad point 1C to determine whether the color channel of each pixel is a bad point channel on the digital image generated by the digital camera device after shooting a scene. . 5. If the request item 4 rides green, the intensity value of the corresponding color channel of one pixel in the -_week _ plural pixels is substituted for the intensity value of the bad point channel. 6. If the value of the corresponding color channel of the pixel of the 4th job of the 4th job is interpolated, replace the intensity value of the channel of the bad point, and the interpolation is the intensity value of each corresponding color channel. , a value between the maximum and minimum intensity values. 7. The method of claim 6, wherein the interpolation is an average of intensity values of corresponding color channels of a plurality of pixels adjacent to the bad point coordinates. 8·—a digital camera device with a dead-end automatic correction function, including a photosensitive element, which converts a (10)-filament image into a touch image to obtain a three-color channel for each pixel in the digital county. Intensity value, the two-color channel is a red channel, a green channel, and a blue-optical device for concentrating on the photosensitive element; - a memory for storing digital data, the digital data containing dead pixel coordinates And the bad point channel; - the bad point processing software 'the bad point processing soft _ root 掳 存在 存在 存在 该 该 该 该 该 该 该 该 该 该 数 数 数 数 数 数 数 数 数 数 数 数 数 数 数 数 数 数 数 数 数 数 数 数 数 数 数 数 数 数 数Whether each color channel of each pixel is a bad point channel, and the intensity value of a bad point channel is replaced by one of the intensity values of the corresponding color channels of the plurality of pixels around the dead point, and the interpolation is replaced by Among the intensity values of the corresponding color channels, a value between the maximum and minimum intensity values; an arithmetic processing unit that performs data operations and data according to the instructions of the bad point processing software Li; and * ',' screen page shown, the digital image system to present after the singular points of the generated correction processing software. The digital camera device of claim 8, wherein the dead point processes the intensity values of the three color channels of each pixel on the digital image of the software, and calculates the average value and standard deviation of the intensity values of the color channels. And calculating a three color channel_value of the digital image, wherein a threshold of each color channel is equal to a standard deviation of the intensity value of the color channel multiplied by a multiple, and each pixel of the digital image is counted Whether the difference between the intensity values of the color channels and the intensity values of the color channels of each pixel is greater than the threshold of each color channel, when the intensity values of the color channels of one pixel are found If the difference is greater than the threshold of the color channel, then the pixel is a dead pixel. The coordinates of the pixel are marked as a bad point coordinate. The color channel is leaked as a bad track, and each of the threshold, the bad point coordinates, and the bad The dot channel is stored in this memory. twenty one