Stripe noise detection method and device, electronic equipment and storage medium

Abstract

The invention relates to a stripe noise detection method and device, electronic equipment and a storage medium. The stripe noise detection method comprises the steps that a to-be-detected image output by a target object and a channel image group corresponding to the to-be-detected image are determined, and channel color information of channel images in the channel image group is different; a target channel image is selected from the channel image group, an image amplitude spectrum corresponding to the target channel image is determined, column pixel information and row pixel information corresponding to the target channel image are determined according to the image amplitude spectrum, and the pixel directions of the column pixel information and the row pixel information are different; performing noise detection on the target channel image according to a preset detection strategy based on the column pixel information and the row pixel information, and determining a noise detection result of the target channel image; and generating noise detection information of the to-be-detected image by using the noise detection result. Through the method provided by the invention, the stripe noise in the image can be quickly and accurately identified.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the technical field of image processing, and particularly relates to a stripe noise detection method and device, electronic equipment and storage medium. BACKGROUND

[0002] With the continuous development of image technology, users have higher and higher requirements for image quality, so that manufacturers of products with shooting functions such as mobile phones and cameras pay more and more attention to image output quality. However, in the production process, CMOS (Complementary Metal Oxide Semiconductor) sensors are widely used in product cameras because of their high resolution and low cost. However, the sensor may cause stripe noise due to power supply noise, external electromagnetic interference and other problems. The stripe noise seriously reduces the image quality and has an adverse effect on the extraction and use of subsequent image information, which requires detection of stripe noise in images. However, the current related detection technology has problems in detection accuracy and detection efficiency, so how to quickly and accurately detect stripe noise in images is a problem that needs to be solved at present. SUMMARY

[0003] To overcome the problems in the related art, the present disclosure provides a stripe noise detection method, device, electronic equipment and storage medium.

[0004] According to a first aspect of an embodiment of the present disclosure, a stripe noise detection method is provided, comprising:

[0005] determining a to-be-detected image output by a target object, and a channel image group corresponding to the to-be-detected image, wherein the channel color information of each channel image in the channel image group is different;

[0006] selecting a target channel image in the channel image group, determining an image amplitude spectrum corresponding to the target channel image, and determining column pixel information and row pixel information corresponding to the target channel image according to the image amplitude spectrum, wherein the pixel directions of the column pixel information and the row pixel information are different;

[0007] performing noise detection on the target channel image according to a preset detection strategy based on the column pixel information and the row pixel information, to determine a noise detection result of the target channel image;

[0008] generating noise detection information of the to-be-detected image by using the noise detection result, wherein the noise detection information is used to verify the image output quality of the target object.

[0009] In an implementation manner, the image amplitude spectrum corresponding to the target channel image is determined, comprising:

[0010] performing Fourier transform on the target channel image to obtain frequency domain data of the target channel image;

[0011] determining an initial image amplitude spectrum based on the frequency domain data, and determining a low frequency component in the initial image amplitude spectrum;

[0012] performing shift processing on the low frequency component to obtain an image amplitude spectrum corresponding to the target channel image.

[0013] In an embodiment, the shift processing on the low frequency component to obtain the image amplitude spectrum corresponding to the target channel image comprises:

[0014] performing shift processing on the low frequency component to obtain an intermediate amplitude spectrum;

[0015] determining cutting information corresponding to the intermediate amplitude spectrum according to image attribute information of the target channel image;

[0016] cutting the intermediate amplitude spectrum according to the cutting information to obtain the image amplitude spectrum corresponding to the target channel image.

[0017] In an embodiment, the determining of the column pixel information and the row pixel information corresponding to the target channel image according to the image amplitude spectrum comprises:

[0018] determining a row pixel point set and a column pixel point set corresponding to the image amplitude spectrum, wherein the row pixel point set comprises row pixel point groups each composed of row pixel points in each row of the image amplitude spectrum, and the column pixel point set comprises column pixel point groups each composed of column pixel points in each column of the image amplitude spectrum;

[0019] calculating a row pixel projection value corresponding to each row pixel point group, and generating the row pixel information corresponding to the target channel image according to each row pixel projection value, and calculating a column pixel projection value corresponding to each column pixel point group, and generating the column pixel information corresponding to the target channel image according to each column pixel projection value.

[0020] In an embodiment, the sub-row pixel information corresponding to any one row pixel point group is determined in the following manner:

[0021] determining a target row pixel point group and a row pixel value of each row pixel point in the target row pixel point group;

[0022] calculating a row pixel projection value of the target row pixel point group based on the row pixel value of each row pixel point;

[0023] generating the row pixel information corresponding to the target channel image according to each row pixel projection value comprises:

[0024] sorting each row pixel projection value in descending order, and selecting a preset number of row pixel projection values as target row pixel projection values according to the sorting result;

[0025] According to the row pixel projection value of each row pixel point group, a row pixel mean value and a row pixel standard deviation value are calculated;

[0026] The target row pixel projection value, the row pixel mean value and the row pixel standard deviation value are taken as row pixel information of the target channel image.

[0027] In an embodiment, the sub-column pixel information corresponding to any one column pixel point group is determined in the following manner:

[0028] A target column pixel point group and a column pixel value of each column pixel point in the target column pixel point group are determined;

[0029] A column pixel projection value of the target column pixel point group is calculated based on the column pixel value;

[0030] The column pixel information corresponding to the target channel image is generated according to each column pixel projection value, including:

[0031] Each column pixel projection value is sorted in descending order, and a preset number of column pixel projection values are selected as target column pixel projection values according to the sorting result;

[0032] According to each column pixel projection value, a column pixel mean value and a column pixel standard deviation value are calculated;

[0033] The target column pixel projection value, the column pixel mean value and the column pixel standard deviation value are taken as column pixel information of the target channel image.

[0034] In an embodiment, the target channel image is subjected to noise detection according to a preset detection strategy based on the column pixel information and the row pixel information, including:

[0035] An initial noise detection data template is determined according to the preset detection strategy, and the initial noise detection data template is updated based on the column pixel information and the row pixel information respectively to obtain column noise detection data and row noise detection data;

[0036] The target channel image is subjected to noise detection according to the column noise detection data and the row noise detection data to obtain a noise detection result of the target channel image.

[0037] In an embodiment, before the noise detection result of the group of channel images is generated, the method further includes:

[0038] An initial noise removal data template is determined according to the preset detection strategy, and the initial noise removal data template is updated based on the column pixel information and the row pixel information respectively to obtain column noise removal data and row noise removal data;

[0039] According to the column noise elimination data and the row noise elimination data, the noise detection result is eliminated, and the eliminated noise detection result is taken as the noise detection result of the target channel image.

[0040] In an implementation, the method further includes:

[0041] In response to receiving the quality inspection instruction for the target object, the inspection environment information is determined.

[0042] According to the inspection environment information, the target object outputs an image to obtain the to-be-detected image.

[0043] The method further includes:

[0044] According to the channel image corresponding to each channel color information, the channel image group corresponding to the to-be-detected image is constructed.

[0045] In an implementation, the method further includes:

[0046] In a case where it is determined that the target object does not meet the object use condition based on the noise detection information, the object attribute information of the target object is acquired.

[0047] Based on the noise detection information and the object attribute information, the inspection information is generated and fed back.

[0048] According to a second aspect of the embodiments of the present disclosure, a stripe noise detection device is provided, including:

[0049] A determination unit is configured to determine a to-be-detected image output by a target object, and a channel image group corresponding to the to-be-detected image, wherein channel color information of each channel image in the channel image group is different.

[0050] A selection unit is configured to select a target channel image from the channel image group, determine an image amplitude spectrum corresponding to the target channel image, and determine column pixel information and row pixel information corresponding to the target channel image according to the image amplitude spectrum, wherein the pixel directions of the column pixel information and the row pixel information are different.

[0051] A detection unit is configured to perform noise detection on the target channel image according to a preset detection strategy based on the column pixel information and the row pixel information, and determine a noise detection result of the target channel image.

[0052] A generation unit is configured to generate noise detection information of the to-be-detected image by using the noise detection result, wherein the noise detection information is used to inspect the image output quality of the target object.

[0053] In an embodiment, the selecting unit determines the image amplitude spectrum corresponding to the target channel image by:

[0054] performing Fourier transform on the target channel image to obtain frequency domain data of the target channel image;

[0055] determining an initial image amplitude spectrum based on the frequency domain data, and determining a low frequency component in the initial image amplitude spectrum;

[0056] performing shift processing on the low frequency component to obtain the image amplitude spectrum corresponding to the target channel image.

[0057] In an embodiment, the selecting unit performs shift processing on the low frequency component to obtain the image amplitude spectrum of the target channel image by:

[0058] performing shift processing on the low frequency component to obtain an intermediate amplitude spectrum;

[0059] determining cutting information corresponding to the intermediate amplitude spectrum according to image attribute information of the target channel image;

[0060] performing cutting processing on the intermediate amplitude spectrum according to the cutting information to obtain the image amplitude spectrum corresponding to the target channel image.

[0061] In an embodiment, the selecting unit determines the column pixel information and the row pixel information corresponding to the target channel image according to the image amplitude spectrum by:

[0062] determining a row pixel point set and a column pixel point set corresponding to the image amplitude spectrum, wherein the row pixel point set comprises row pixel point groups each consisting of row pixel points in each row of the image amplitude spectrum, and the column pixel point set comprises column pixel point groups each consisting of column pixel points in each column of the image amplitude spectrum;

[0063] calculating a row pixel projection value corresponding to each row pixel point group, and generating the row pixel information corresponding to the target channel image according to each row pixel projection value, and calculating a column pixel projection value corresponding to each column pixel point group, and generating the column pixel information corresponding to the target channel image according to each column pixel projection value.

[0064] In an embodiment, the selecting unit determines the sub-row pixel information corresponding to any one row pixel point group by:

[0065] determining a target row pixel point group, and a row pixel value of each row pixel point in the target row pixel point group;

[0066] calculating a row pixel projection value of the target row pixel point group based on the row pixel value of each row pixel point;

[0067] Generating row pixel information corresponding to the target channel image according to each row pixel projection value, comprising:

[0068] Sorting each row pixel projection value in descending order, and selecting a preset number of row pixel projection values as target row pixel projection values according to the sorting result;

[0069] According to the row pixel projection value of each row pixel point group, the row pixel mean and the row pixel standard deviation value are calculated;

[0070] The target row pixel projection value, the row pixel mean and the row pixel standard deviation value are used as the row pixel information of the target channel image.

[0071] In an embodiment, the selection unit determines the sub-column pixel information corresponding to any one column pixel point group by the following method, comprising:

[0072] Determine the target column pixel point group, and the column pixel value of each column pixel point in the target column pixel point group;

[0073] Calculate the column pixel projection value of the target column pixel point group based on the column pixel value;

[0074] Generating column pixel information corresponding to the target channel image according to each column pixel projection value, comprising:

[0075] Sorting each column pixel projection value in descending order, and selecting a preset number of column pixel projection values as target column pixel projection values according to the sorting result;

[0076] According to each column pixel projection value, the column pixel mean and the column pixel standard deviation value are calculated;

[0077] The target column pixel projection value, the column pixel mean and the column pixel standard deviation value are used as the column pixel information of the target channel image.

[0078] In an embodiment, the target channel image is detected for noise according to a preset detection strategy based on the column pixel information and the row pixel information by the following method, comprising:

[0079] According to the preset detection strategy, an initial noise detection data template is determined, and the initial noise detection data template is updated based on the column pixel information and the row pixel information respectively to obtain column noise detection data and row noise detection data;

[0080] According to the column noise detection data and the row noise detection data, the target channel image is detected for noise to obtain the noise detection result of the target channel image.

[0081] In an implementation manner, the detection unit is further configured to determine an initial noise-removed data template according to a preset detection strategy, update the initial noise-removed data template based on the column pixel information and the row pixel information respectively to obtain column noise-removed data and row noise-removed data, and perform noise removal on the noise detection result according to the column noise-removed data and the row noise-removed data, and take the noise-removed noise detection result as the noise detection result of the target channel image.

[0082] In an implementation manner, the determination unit determines the to-be-detected image output by the target object and the channel image group corresponding to the to-be-detected image in the following manner:

[0083] In response to receiving a quality inspection instruction for the target object, the determination unit determines inspection environment information.

[0084] The target object outputs an image according to the inspection environment information to obtain the to-be-detected image.

[0085] The determination unit determines a plurality of channel color information corresponding to the to-be-detected image, and analyzes the to-be-detected image according to the plurality of channel color information to obtain a channel image corresponding to each channel color information.

[0086] The determination unit constructs a channel image group corresponding to the to-be-detected image according to the channel image corresponding to each channel color information.

[0087] In an implementation manner, the apparatus further includes a feedback unit configured to acquire object attribute information of the target object in a case where it is determined that the target object does not meet the object use condition based on the noise detection information.

[0088] The feedback unit generates inspection information based on the noise detection information and the object attribute information and feeds back the inspection information.

[0089] According to a third aspect of the embodiments of the present disclosure, an electronic device is provided, including: a processor; a memory for storing processor-executable instructions; and wherein the processor is configured to execute the stripe noise method in the first aspect or any one of the implementation manners of the first aspect.

[0090] According to a fourth aspect of the embodiments of the present disclosure, a storage medium is provided, and the storage medium stores instructions, when the instructions in the storage medium are executed by a processor, the processor can execute the stripe noise method in the first aspect or any one of the implementation manners of the first aspect.

[0091] The technical scheme provided by the embodiment of the present disclosure can have the following beneficial effects: the target object outputted image to be detected and the channel image group corresponding to the image to be detected are determined, the image to be detected is split according to the channel color information, which facilitates subsequent image detection on different channel images of the image to be detected, and improves the image detection efficiency. The target channel image is selected from the plurality of channel images, the image amplitude spectrum of the target channel image is determined, the column pixel information and the row pixel information are determined according to the image amplitude spectrum, and the noise detection of the target channel image is performed based on the column pixel information and the row pixel information according to the preset detection strategy, and the noise detection result of the channel image group is determined. The image noise detection based on the frequency spectrum angle of the image amplitude spectrum is realized, the noise detection of the target channel image can be completed according to the preset detection strategy by using the column pixel information and the row pixel information subsequently, the detection complexity is reduced, and the detection efficiency is improved. Finally, the noise detection information of the image to be detected is generated by using the noise detection result of the channel image group, so that the noise detection information can more comprehensively reflect the image quality of the image to be detected, and reliable and effective inspection data is provided when the image output quality of the target object is inspected subsequently by using the noise detection information.

[0092] It should be understood that the foregoing general description and the following detailed description are only exemplary and explanatory, and are not limiting to the present disclosure. BRIEF DESCRIPTION OF DRAWINGS

[0093] The accompanying drawings, which are incorporated into and form a part of the specification, illustrate one embodiment consistent with the present disclosure and, together with the description, serve to explain the principles of the disclosure.

[0094] Figure 1 A flowchart of a stripe noise detection method according to an embodiment of the present disclosure is shown.

[0095] Figure 2 A schematic diagram of an image amplitude spectrum in a stripe noise detection method according to an embodiment of the present disclosure is shown.

[0096] Figure 3 A flowchart of determining an image to be detected and a channel image group in a stripe noise detection method according to an embodiment of the present disclosure is shown.

[0097] Figure 4 A schematic diagram of an image to be detected in a stripe noise detection method according to an embodiment of the present disclosure is shown.

[0098] Figure 5 A flowchart of determining an image amplitude spectrum in a stripe noise detection method according to an embodiment of the present disclosure is shown.

[0099] Figure 6A flowchart of calculating an image amplitude spectrum in a method for detecting stripe noise is shown.

[0100] Figure 7 A flowchart of determining column pixel information and row pixel information in a method for detecting stripe noise is shown.

[0101] Figure 8 A flowchart of determining row pixel projection values in a method for detecting stripe noise is shown.

[0102] Figure 9 A schematic diagram of row pixel projection values in a method for detecting stripe noise is shown.

[0103] Figure 10 A flowchart of determining row pixel projection values in a method for detecting stripe noise is shown.

[0104] Figure 11 A flowchart of determining column pixel projection values in a method for detecting stripe noise is shown.

[0105] Figure 12 A schematic diagram of column pixel projection values in a method for detecting stripe noise is shown.

[0106] Figure 13 A flowchart of determining column pixel information in a method for detecting stripe noise is shown.

[0107] Figure 14 A flowchart of noise detection in a method for detecting stripe noise is shown.

[0108] Figure 15 A flowchart of noise intensity detection in a method for detecting stripe noise is shown.

[0109] Figure 16 A flowchart of information feedback in a method for detecting stripe noise is shown.

[0110] Figure 17 A flowchart of a processing procedure of a method for detecting stripe noise is shown.

[0111] Figure 18 A block diagram of a device for detecting stripe noise is shown.

[0112] Figure 19A block diagram of an apparatus for stripe noise detection is shown according to an embodiment of the present disclosure.

[0113] Figure 20 A block diagram of an apparatus for stripe noise detection is shown according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0114] The exemplary embodiments will be described in detail herein with reference to the attached drawings. The following description is with reference to the drawings, in which like numerals represent like elements, unless otherwise described. The following exemplary embodiments described herein are not meant to represent all embodiments consistent with the present disclosure.

[0115] First, the nomenclature related to one or more embodiments of the present disclosure is explained.

[0116] Fixed pattern noise: Fixed pattern noise (FPN) is divided into row FPN and column FPN according to the FPN noise formation mechanism. This noise shows a deviation in brightness or color at a fixed position in an image, which seriously affects the quality of the image.

[0117] CMOS sensor: CMOS (Complementary Metal Oxide Semiconductor) sensors can be used as image sensors. Currently, the photosensitive elements of common digital products are mainly CMOS sensors.

[0118] At present, CMOS sensors are widely used in mobile phone cameras due to their high resolution and low cost. For CMOS sensors, each pixel has its own corresponding output amplifier, which may cause additional non-uniformity and vertical FPN due to vertical amplifiers. Stripe noise is a common phenomenon in mobile phone imaging, which seriously reduces the image quality and has an adverse effect on the extraction and use of subsequent image information. Sensor power noise, external electromagnetic interference, OIS frequency interference, and response differences of the sensor to the signal response area may all cause stripe noise. Stripe noise exhibits different characteristics from other general random noise in the image, and the brightness of a specific column or row of the image is darker or brighter than the adjacent rows or columns. Existing stripe noise detection methods are based on the periodicity of stripes or the sharp changes in the average value, standard deviation, gradient, and other features of the brightness of adjacent positions in the image. However, when using these shallow features for evaluation, it is challenging to address the irregular distribution of stripe noise with small neighborhood changes. In recent methods, deep features of stripe noise such as structural features and directional features are extracted and comprehensively judged, achieving good performance in detection accuracy. However, due to the large size of the image and the diversity of images from different sensors, on the one hand, the calculation speed of extracting artificially defined features for judgment is slow, such as calculating the mean and variance of the entire image for subsequent processing. On the other hand, these features may not be suitable for noise detection in multi-sensor images.

[0119] Therefore, the present disclosure provides a stripe noise detection method for quickly and accurately detecting stripe noise in an image, facilitating the inspection of the image output quality of a sensor. The present disclosure also relates to a stripe noise detection device, an electronic device, and a storage medium.

[0120] Figure 1 A flowchart of a stripe noise detection method according to an embodiment of the present disclosure is shown in FIG. 1, which includes steps S102-S108. Figure 1

[0121] In step S102, a target object outputs a to-be-detected image, and a channel image group corresponding to the to-be-detected image is determined, wherein the channel color information of each channel image in the channel image group is different.

[0122] In step S104, a target channel image is selected from the channel image group, an image amplitude spectrum corresponding to the target channel image is determined, and column pixel information and row pixel information corresponding to the target channel image are determined according to the image amplitude spectrum, wherein the pixel directions of the column pixel information and the row pixel information are different.

[0123] ​In step S106, based on the column pixel information and the row pixel information, the target channel image is detected according to a preset detection strategy, and a noise detection result of the target channel image is determined.

[0124] In step S108, the noise detection result is used to generate noise detection information of the image to be detected, wherein the noise detection information is used to verify the image output quality of the target object.

[0125] In the embodiments of the present disclosure, the target object can be understood as an object outputting the image to be detected, and the image to be detected is an image that needs to be detected for stripe noise. In actual applications, the target object can be a sensor used in a data product, such as a CMOS sensor of a mobile phone camera. The stripe noise detection method provided by the present disclosure can be applied to verify the image output quality of the sensor, or can be applied to detect the image noise of only the multiple images output by the sensor. The specific application can be determined according to actual conditions.

[0126] In practice, when the image to be detected output by the target object is determined, that is, the image data output by the sensor is obtained, since the mainstream CMOS sensor outputs RAW image data in Bayer mosaic format, this data format cannot be directly viewed and must be converted into a common RGB or YUV format to be supported by mainstream image processing software. For camera products, the RGB or YUV format image needs to be further converted into a JPEG format image for easy storage. Therefore, when detecting the stripe noise of the image output by the target object, it is necessary to consider detecting the RAW image data in order to more truly and accurately identify the noise in the image, and to avoid the situation of missing noise caused by the change of image format. The stripe noise detection method provided by the present disclosure can quickly detect the stripe noise of the RAW format image data, and accordingly, the JPEG format image obtained by converting the RAW format image can also use the method.

[0127] In addition, in order to consider the influence of each color channel in the image on noise and improve the accuracy of stripe noise detection, when performing stripe noise detection on the to-be-detected image, detection can be performed on the to-be-detected image at different channel levels, and the final detection result of the entire to-be-detected image is merged according to the detection results of different channel images. Therefore, it is necessary to determine the channel image group corresponding to the to-be-detected image. The channel image group can be understood as a combination of channel images corresponding to different color channels of the to-be-detected image, that is, the channel color information of each channel image in the channel image group is different. The channel color information is determined according to the image format of the to-be-detected image. For example, an image in RAW format can be split into four different channel images R, Gr, Gb, and B. In the embodiment of the present disclosure, the channel color information represented by R is red, the channel color information represented by Gr and GB is two different greens, and the channel color information represented by B is blue.

[0128] In the embodiment of the present disclosure, the target channel image can be understood as a channel image selected from the channel image group. Since the channel image group of the to-be-detected image includes multiple different channel images, when performing stripe noise detection, it is necessary to detect them one by one. The channel image can be selected as the target channel image in sequence from the channel image group, and the detection is performed in sequence. The channel image can also be selected as the target channel image in sequence from the channel image group, and the detection is performed in parallel. In the above two methods, the detection method of each channel image is the same, that is, the image amplitude spectrum corresponding to the target channel image needs to be determined first, and then the column pixel information and the row pixel information corresponding to the target channel image are determined according to the image amplitude spectrum.

[0129] In actual application, in order to improve the detection speed, when performing image noise detection, image noise detection can be performed from the frequency domain. Therefore, the image amplitude spectrum corresponding to the target channel image can be determined, and then the column pixel information and the row pixel information corresponding to the target channel image are determined based on the image amplitude spectrum. The image amplitude spectrum is used to observe the pixel amplitude in the target channel image, so that the noise in the image can be effectively determined subsequently. Referring to Figure 2 , Figure 2 A schematic diagram of an image amplitude spectrum in a stripe noise detection method according to an embodiment of the present disclosure is shown. The amplitudes of each pixel point can be observed in the image amplitude spectrum, so that the stripe noise with high noise intensity can be clearly displayed. In order to realize accurate detection of stripe noise, the column pixel information and the row pixel information of the target channel image need to be determined based on the image amplitude spectrum. In the embodiment of the present disclosure, the pixel directions of the column pixel information and the row pixel information are different, that is, the column pixel information is determined by the pixel points in the Y-axis direction of the image, and the row pixel information is determined by the pixel points in the X-axis direction of the image. The column pixel information and the row pixel information can be used to accurately detect the stripe noise in the image.

[0130] In this embodiment of the disclosure, the preset detection strategy can be understood as a strategy for detecting stripe noise. The preset detection strategy may include a detection method for detecting stripe noise. By following the preset detection strategy, stripe noise in the target channel image can be quickly and accurately determined, thereby determining the noise detection result of the channel image group.

[0131] In this embodiment of the disclosure, the noise detection result may include the detection result corresponding to each channel image. Based on the detection result corresponding to each channel image, noise detection information of the final image to be detected can be generated. The noise detection information of the image to be detected may be presented to the user in text form or in image form.

[0132] In practical applications, if the noise detection information is in text form, it can simply show whether noise is present or absent in the current image. If the noise detection information is in image form, the projection map of the channel image can be displayed so that the user can visually see the stripe noise.

[0133] In one embodiment of this disclosure, it is necessary to perform anomaly detection on the sensor module, that is, to filter out sensors with problematic image output. During detection, the image to be detected output by the sensor is first determined; this image can be RAW format image data. To verify the image output quality of the sensor, it can be determined by detecting stripe noise in the image output by the sensor. When performing stripe noise detection on the image to be detected, it is necessary to split the image into channel images to obtain a group of channel images corresponding to the image to be detected. Each group of channel images corresponds to four different channel images: R, Gr, Gb, and B.

[0134] Furthermore, to more accurately detect stripe noise in an image, the final detection result can be determined by separately detecting channel images with different color channels. In one implementation, the target object output image to be detected, and the corresponding group of channel images, can be determined by... Figure 3 Steps S12-S16 are implemented. Figure 3 A flowchart illustrating the determination of an image to be detected and a group of channel patterns in a stripe noise detection method according to an embodiment of the present disclosure is shown.

[0135] In step S12, in response to receiving a quality inspection instruction for the target object, inspection environment information is determined.

[0136] In step S14, the target object outputs an image based on the inspection environment information to obtain the image to be inspected.

[0137] In step S16, multiple channel color information corresponding to the to-be-detected image is determined, and the to-be-detected image is parsed according to the multiple channel color information to obtain a channel image corresponding to each channel color information. According to the channel image corresponding to each channel color information, a channel image group corresponding to the to-be-detected image is constructed.

[0138] The quality inspection instruction can be understood as an inspection of the image output quality of the sensor. The image output quality of the sensor can be determined to determine whether the sensor has a problem, so that the problem sensor module can be filtered out in the image output stage of the sensor. The inspection environment information can be understood as the image output environment of the sensor during the inspection of the sensor. Since the image output quality of the sensor is inspected according to the image output by the sensor, the inspection needs to be performed in a specific inspection environment. In actual application, in order to accurately inspect the image output quality of the sensor, the image output environment of the sensor needs to be set, which is determined by the inspection environment information. Generally, the inspection environment information is a light source-free environment, that is, the sensor is required to output an image in the absence of a light source. See Figure 4 , Figure 4 An image schematic diagram in a stripe noise detection method according to an embodiment of the present disclosure is shown. The image is a RAW format image directly output by the sensor. Since the image is output in a light source-free environment, the image does not have any light or other content, which facilitates subsequent stripe noise detection based on the image. Therefore, after the target object outputs an image according to the inspection environment information, an image that can be detected is obtained as a to-be-detected image. It should be noted that the inspection environment information can also be other environments, for example, the stripe noise is inspected in the presence of a light source.

[0139] In implementation, in order to ensure that a more accurate and reliable image detection result is obtained, the to-be-detected image can be channel split to obtain multiple different channel images. Subsequently, the multiple different channel images are detected for stripe noise, thereby improving detection accuracy and detection efficiency. The multiple channel color information corresponding to the to-be-detected image can be understood as image channel color corresponding to the to-be-detected image, for example, R channel, G channel, and B channel in an RGB color space. The R channel represents a red channel, the G channel represents a green channel, and the B channel represents a blue channel. After the to-be-detected image is parsed by different channel color information, a channel image corresponding to each channel color information is obtained. According to each channel image, a channel image group corresponding to the to-be-detected image is constructed, which facilitates subsequent stripe noise detection of each channel component in the channel image group. Accordingly, a parallel detection mode can be selected to improve detection efficiency.

[0140] In an embodiment of the present disclosure, a quality inspection instruction for a target sensor is received, it is determined that the inspection environment information is a light-free environment, the target sensor outputs an image in the light-free environment, the output image data is taken as a to-be-detected image, the to-be-detected image is parsed, and is split into four channels of R, Gr, Gb and B, a channel image corresponding to each channel is obtained, and a channel image group is formed.

[0141] Therefore, when the sensor needs to be inspected, the sensor outputs an image in a fixed environment, the accuracy of subsequent image noise detection is improved, and in the detection process, the to-be-detected image is split into multiple different channel images, the multiple different channel images are detected, the accuracy of image noise detection is improved, and the detection efficiency is also improved.

[0142] In an embodiment of the present disclosure, referring to the above example, a channel image of the R channel is selected as a target channel image from the channel image group containing four different channel images, an image amplitude spectrum is determined based on each pixel point in the target channel image, column pixel information and row pixel information in the target channel image are determined based on the image amplitude spectrum, and stripe noise detection for the target channel image can be implemented based on the column pixel information and the row pixel information.

[0143] Further, in order to more intuitively view the positions of the low-frequency component and the high-frequency component from the image amplitude spectrum, the low-frequency component in the image amplitude spectrum can be shifted. In an embodiment, the image amplitude spectrum corresponding to the target channel image can be obtained by Figure 5 steps S22-S24, Figure 5 A flowchart for determining an image amplitude spectrum in a stripe noise detection method according to an embodiment of the present disclosure is shown.

[0144] In step S22, Fourier transform is performed on the target channel image to obtain frequency domain data of the target channel image.

[0145] In step S24, the low-frequency component in the frequency domain data is shifted to obtain target frequency domain data, and the image amplitude spectrum corresponding to the target channel image is calculated based on the target frequency domain data.

[0146] The frequency domain data can be understood as data obtained after Fourier transform of the target channel image. In actual application, after Fourier transform of the image, the obtained data is a complex matrix, each element of which is composed of a real part and an imaginary part. The complex matrix contains frequency domain information of the image and can be used to analyze various frequency components of the image. In order to more intuitively view the positions of the low-frequency components and the high-frequency components in the frequency domain data, the low-frequency components in the frequency domain data can be moved to the center position of the image spectrum corresponding to the frequency domain data by shifting, so as to obtain target frequency domain data. After obtaining the target frequency domain data of the target channel image, in order to analyze the existence and intensity of the stripe noise in the target channel image, the image amplitude spectrum of the target channel image can also be calculated based on the frequency domain data. The image amplitude spectrum can be understood as amplitude spectrum data obtained by modulus calculation based on the frequency domain data. The amplitude spectrum is usually used to analyze the frequency characteristics of the image in image processing. For stripe noise detection, the amplitude spectrum can help us identify whether there is a stripe pattern of a specific frequency. If the stripe noise exists, it will show obvious peak values in the amplitude spectrum, which correspond to the frequency of the stripe.

[0147] In implementation, a two-dimensional discrete Fourier transform is performed on the target channel image to obtain a complex matrix F, i.e. frequency domain data. In order to more intuitively view the low-frequency components, the low-frequency components can be moved to the center position of the image spectrum corresponding to the frequency domain data by using a shift function, so as to obtain target frequency domain data, i.e. frequency domain data after moving. The frequency domain data contains amplitude spectrum information and phase spectrum information. The amplitude spectrum information is needed in the subsequent stripe detection process, so the image amplitude spectrum |F| can be calculated after modulus calculation of the complex matrix F. The subsequent stripe noise detection is performed according to the calculated image amplitude spectrum |F|.

[0148] In summary, by performing Fourier transform on the target channel image, the frequency domain data of the target channel image in the frequency domain angle is obtained, so that the subsequent stripe noise detection starts from the frequency domain angle, the detection complexity is reduced, the low-frequency components in the frequency domain data are shifted, which is convenient for subsequent observation and improves the detection accuracy. The image amplitude spectrum is calculated based on the frequency domain data, which is a key index for evaluating the energy of different frequency components in the image, so that the stripe noise in the image can be more conveniently and accurately detected subsequently.

[0149] Further, in the process of calculating the image amplitude spectrum based on the target frequency domain data, in order to reduce the calculation amount of the subsequent stripe noise, the final image amplitude spectrum can be determined by center cutting. In one implementation, the image amplitude spectrum corresponding to the target channel image can be calculated based on the target frequency domain data by steps S32-S34 in Figure 6 Figure 6 ​A flow chart of calculating an image amplitude spectrum in a method for detecting stripe noise is shown according to an embodiment of the present disclosure.

[0150] In step S32, an initial image amplitude spectrum corresponding to the target channel image is calculated according to the target frequency domain data, and the clipping information of the initial image amplitude spectrum is determined according to the image attribute information of the target channel image.

[0151] In step S34, the initial image amplitude spectrum is clipped according to the clipping information to obtain the image amplitude spectrum corresponding to the target channel image.

[0152] The target frequency domain data is the frequency domain data obtained after Fourier transform and low-frequency component shift of the image in the previous step. The amplitude spectrum of the image can be calculated using the target frequency domain data. The initial image amplitude spectrum corresponding to the target channel image is obtained at this time. In order to reduce the calculation amount of subsequent stripe noise detection, the center of the initial image amplitude spectrum can be selected as the final image amplitude spectrum. In order to clip the image amplitude spectrum in a suitable range, the clipping information needs to be determined according to the image attribute information of the target channel image.

[0153] In actual application, the image attribute information of the target channel image is the length and width of the target channel image. The clipping information is calculated by the length and width of the target channel image, and the clipping information is the length of the clipped image. The center of the initial image amplitude spectrum is taken as the center during clipping, and the clipping is performed according to the clipping information. The calculation method of the corresponding image attribute information and the clipping information can be referred to formula (1), which is as follows:

[0154]

[0155] Wherein, w is the clipping length in the clipping information, width is the length of the target channel image, and heigh is the width of the target channel image.

[0156] In summary, after the clipping information is calculated in the above-mentioned manner, the image amplitude spectrum can be clipped in the initial image amplitude spectrum according to the clipping information. The image amplitude spectrum can be regarded as the ROI region in the initial image amplitude spectrum. The ROI (Region of Interest) region is the region that needs to be paid special attention or processed in the image. Subsequently, the image amplitude spectrum is processed, which not only reduces the calculation amount required for stripe noise detection, but also significantly improves the detection efficiency and accuracy.

[0157] Further, in order to accurately identify the image amplitude spectrum, the corresponding column pixel information and row pixel information need to be determined first. In an embodiment, the column pixel information and the row pixel information corresponding to the target channel image are determined according to the image amplitude spectrum, which can be determined by taking the center of the image amplitude spectrum as the center, and the length of the image amplitude spectrum as the length of the column pixel information and the length of the image amplitude spectrum as the length of the row pixel information. Figure 7Steps S42-S44 in the method shown in FIG. 4 are implemented as follows. Figure 7 A flowchart of determining column pixel information and row pixel information in a method for detecting stripe noise is shown.

[0158] In step S42, a set of row pixel points and a set of column pixel points corresponding to the image amplitude spectrum are determined, where the set of row pixel points includes groups of row pixel points each composed of pixel points in a row of the image amplitude spectrum, and the set of column pixel points includes groups of column pixel points each composed of pixel points in a column of the image amplitude spectrum.

[0159] In step S44, a row pixel projection value corresponding to each group of row pixel points is calculated, and row pixel information corresponding to the target channel image is generated according to each row pixel projection value; and a column pixel projection value corresponding to each group of column pixel points is calculated, and column pixel information corresponding to the target channel image is generated according to each column pixel projection value.

[0160] The set of row pixel points of the image amplitude spectrum can be understood as a set composed of groups of row pixel points each composed of pixel points in a row of the image amplitude spectrum. The set of column pixel points can be understood as a set composed of groups of column pixel points each composed of pixel points in a column of the image amplitude spectrum. For example, all pixel points in the first row of the image amplitude spectrum can form a first row of pixel points, all pixel points in the second row can form a second row of pixel points, and so on. Each group of row pixel points corresponding to a row is used to form the set of row pixel points. Similarly, each group of column pixel points corresponding to a column is used to form the set of column pixel points.

[0161] In actual applications, after each group of row pixel points is determined, a row pixel projection value of each group of row pixel points can be calculated. The row pixel projection value can be understood as being calculated from projection values of each pixel point in the group of row pixel points. The row pixel projection value is an average value calculated from row projection values of each pixel point in each group of row pixel points. Similarly, the column pixel projection value is an average value calculated from column projection values of each pixel point in each group of column pixel points. After the row pixel projection value of each row pixel point is calculated, row pixel information of the target channel image can be generated according to each row pixel projection value. The row pixel information of the target channel image is calculated from pixel points in the image amplitude spectrum. The row pixel information can include information required for subsequent stripe noise detection in the x-axis direction, such as the maximum value point in the row pixel projection value, the row pixel mean, and the row pixel standard deviation. Correspondingly, the column pixel information can include information required for subsequent stripe noise detection in the y-axis direction. Similarly, the column pixel information also includes the maximum value point in the column pixel projection value, the column pixel mean, and the column pixel standard deviation.

[0162] In summary, after calculating the row pixel information and the column pixel information corresponding to the target channel image based on the row pixel point set and the column pixel point set in the image amplitude spectrum, the row pixel information and the column pixel information can be used for subsequent detection of the stripe noise, thereby reducing the detection complexity and improving the detection efficiency.

[0163] Further, the determination of the row pixel projection value corresponding to any one row pixel point group can be implemented through steps S52-S54 in Figure 8 Figure 8 FIG. 5 shows a flowchart of determining the row pixel projection value in a stripe noise detection method according to an embodiment of the present disclosure.

[0164] In step S52, a target row pixel point group and the row pixel value of each row pixel point in the target row pixel point group are determined.

[0165] In step S54, the row pixel projection value of the target row pixel point group is calculated based on the row pixel value of each row pixel point.

[0166] In the image amplitude spectrum, the row pixel point set includes a plurality of row pixel point groups. In this embodiment, the calculation of the row pixel projection value of one row pixel point group is described. First, one row pixel point group is selected as the target row pixel point group from the plurality of row pixel point groups. The row pixel value of each row pixel point in the target row pixel point group is determined. The row pixel value of the row pixel point can be the value corresponding to the pixel point in the image amplitude spectrum, which is generally the brightness value of the pixel point. After obtaining the row pixel values of all the row pixel points in the row, the mean value of the row pixel values of all the row pixel points in the row is calculated, thereby calculating the row pixel projection value. Similarly, the row pixel projection value corresponding to each row pixel point group can be calculated according to the method. After obtaining each row pixel projection value, the row pixel projection value can be used for projection, thereby generating a visual projection diagram. See Figure 9 Figure 9 FIG. 5 shows a schematic diagram of the row pixel projection value in a stripe noise detection method according to an embodiment of the present disclosure. As can be seen from FIG. 5, the continuous row pixel projection value is shown in the x-axis direction in Figure 9

[0167] Correspondingly, the generation of the row pixel information corresponding to the target channel image according to each row pixel projection value can be implemented through steps S62-S64 in Figure 10 Figure 10 FIG. 6 shows a flowchart of determining the row pixel projection value in a stripe noise detection method according to an embodiment of the present disclosure.

[0168] ​​​​In step S62, each row pixel projection value is sorted in descending order, and a preset number of row pixel projection values are selected as target row pixel projection values according to the sorting result.

[0169] In step S64, the row pixel mean value and the row pixel standard deviation value are calculated according to the row pixel projection value of each row pixel point group. The target row pixel projection value, the row pixel mean value and the row pixel standard deviation value are taken as the row pixel information of the target channel image.

[0170] The target row pixel projection value can be understood as a row pixel projection value selected from all row pixel projection values. In the stripe noise detection method provided in the present disclosure, the selection of the target row pixel projection value can be determined according to the size of the row pixel projection value, because the subsequent detection method involves using the row pixel projection value to determine whether the row has noise. Therefore, the row pixel projection value needs to be selected according to the size of the row pixel projection value.

[0171] In actual application, each row pixel projection value can be sorted in descending order first, and then a preset number of row pixel projection values are selected as target row pixel projection values. In an embodiment of the present disclosure, the first maximum value point and the second maximum value point of the row pixel projection value are selected as the target row pixel projection value. Then, the row pixel mean value is calculated according to the row pixel projection value of each row pixel point, that is, the mean value is calculated by accumulation. Finally, the row pixel standard deviation value is calculated, which is the arithmetic square root of the variance calculated by using each row pixel projection value. Finally, the first maximum value point, the second maximum value point, the mean value and the standard deviation are taken as the row pixel information.

[0172] Further, the determination of the column pixel projection value corresponding to any one column pixel point group can be realized by steps S72-S74 in Figure 11 , Figure 11 A flow chart for determining the column pixel projection value in a stripe noise detection method according to an embodiment of the present disclosure is shown.

[0173] In step S72, a target column pixel point group and the column pixel value of each column pixel point in the target column pixel point group are determined.

[0174] In step S74, the column pixel projection value of the target column pixel point group is calculated based on the column pixel value.

[0175] The target column pixel point group is a column pixel point group selected from a plurality of column pixel point groups. Similarly, when calculating the column pixel projection value of the target column pixel point group, the column pixel value of each column pixel point needs to be accumulated to obtain the mean value, so as to obtain the column pixel projection value. See Figure 12 , Figure 12A schematic diagram of column pixel projection values in a stripe noise detection method according to an embodiment of the present disclosure is shown, and it can be seen that Figure 12 The curve value corresponding to any point in the y-axis direction is the column pixel projection value corresponding to the point.

[0176] Accordingly, the column pixel information corresponding to the target channel image can be generated according to each column pixel projection value, which can be achieved by Figure 13 steps S82-S84 in the method, Figure 13 A flowchart of determining column pixel information in a stripe noise detection method according to an embodiment of the present disclosure is shown.

[0177] In step S82, each column pixel projection value is sorted in descending order, and a preset number of column pixel projection values are selected as target column pixel projection values according to the sorting result.

[0178] In step S84, the column pixel mean and the column pixel standard deviation are calculated according to each column pixel projection value. The target column pixel projection value, the column pixel mean and the column pixel standard deviation are taken as the column pixel information of the target channel image.

[0179] Similarly, the column pixel information is determined in the same way as the row pixel information. First, all column pixel projection values are sorted, and then the maximum value point and the second maximum value point are selected. The column pixel mean and the column pixel standard deviation are calculated and taken as the column pixel information of the target channel image.

[0180] In summary, after the row pixel information and the column pixel information of the target channel image are determined in the above manner, the subsequent stripe noise detection can be performed based on simple calculation, thereby avoiding the problem of low detection efficiency caused by complex calculation.

[0181] Further, in order to quickly and accurately determine the stripe noise in the channel image, the detection data template in the preset detection strategy can be used for implementation. In an embodiment, based on the column pixel information and the row pixel information, the target channel image is subjected to noise detection according to the preset detection strategy, which can be achieved by Figure 14 steps S92-S94 in the method, Figure 14 A flowchart of noise detection in a stripe noise detection method according to an embodiment of the present disclosure is shown.

[0182] In step S92, the initial noise detection data template is determined according to the preset detection strategy, and the initial noise detection data template is updated based on the column pixel information and the row pixel information to obtain column noise detection data and row noise detection data.

[0183] In step S94, noise detection is performed on the target channel image according to the column noise detection data and the row noise detection data, and a noise detection result of the target channel image is obtained.

[0184] The initial noise detection data model can be determined according to a preset detection strategy. The initial noise detection data model can be understood as a noise data model used to detect whether there is stripe noise in the image. In actual application, the initial noise detection data model is a noise detection inequality max1>max2+delta*std, where max1 is a maximum value point obtained from the row pixel information or the column pixel information, max2 is a second maximum value point in the row pixel information or the column pixel information, delta is an adjustable proportional coefficient, and std is a row pixel standard deviation value or a column pixel standard deviation value. By adjusting the size of delta, the test standard for whether the stripe noise exists can be controlled, so that the noise invisible to the naked eye can be accurately screened out.

[0185] In actual application, after the initial noise detection data model is obtained, that is, after the noise detection inequality is determined, the row pixel information or the column pixel information can be respectively brought into the noise detection inequality. If the inequality is established, it indicates that there is noise in the current target channel image. In implementation, the channel image of each channel can be detected in the above manner, so as to obtain the detection result of each channel image, that is, the detection result of the channel image group. Subsequently, the detection information of the image to be detected can be generated according to the detection result of the channel image group.

[0186] Further, in order to avoid that some stripe noise with weak noise intensity is detected and the sensor is unreasonably tested to have a problem, the stripe noise with weak noise intensity can be removed. After the noise detection result of the channel image group is generated, the noise intensity detection can be implemented through steps S1002-S1004 in FIG. 10. Figure 15 Figure 15 FIG. 10 shows a flowchart of noise intensity detection in a stripe noise detection method according to an embodiment of the present disclosure.

[0187] In step S1002, an initial noise removal data model is determined according to a preset detection strategy, and the initial noise removal data model is updated based on the column pixel information and the row pixel information respectively, to obtain column noise removal data and row noise removal data.

[0188] In step S1004, noise removal is performed on the noise detection result according to the column noise removal data and the row noise removal data, and the noise detection result after removal is taken as the noise detection result of the target channel image.

[0189] ​The initial noise-removed data template can be understood as a data template used to remove data with weak noise intensity. In actual application, the initial noise-removed data template can be a noise-removed inequality wherein max1 is the maximum value point, avg is the row pixel mean value or the column pixel mean value, and k is an adjustable proportion coefficient. By adjusting the value of k, the intensity of the stripe noise can be controlled. After detecting that the target channel image has noise according to the above method, the noise-removed inequality can be used to further filter the noise intensity. If the noise is weak, the noise can be removed, and it is considered that the target channel image does not have stripe noise, thereby avoiding unreasonable problem inspection of the sensor.

[0190] In summary, after the existence and intensity of the stripe noise are screened and detected, the noise detection result of the current target channel image can be obtained, and the noise detection result of each channel image can be used to generate noise detection information of the final image to be detected.

[0191] In implementation, after the noise detection information of the image to be detected is determined, feedback of the noise detection information can be performed. The information feedback can be achieved by Figure 16 steps S1102-S1104, Figure 16 FIG. 8 shows a flowchart of information feedback in a stripe noise detection method according to an embodiment of the present disclosure.

[0192] In step S1102, in a case where it is determined that the target object does not satisfy the object use condition based on the noise detection information, object attribute information of the target object is obtained.

[0193] In step S1104, inspection information is generated based on the noise detection information and the object attribute information, and the inspection information is fed back.

[0194] The target object is the sensor outputting the image to be detected, the object use condition can be understood as a use condition of the sensor, and the target object satisfying the object use condition can be understood as the sensor meeting the use condition. Therefore, the object use condition can be that there is no noise in the image to be detected. Thus, after the noise detection information of the image to be detected is obtained, whether the current sensor satisfies the use condition can be determined based on the noise detection information.

[0195] In actual application, if it is determined according to the noise detection information that there is no noise in the to-be-detected image, it indicates that the image output quality of the sensor is high and meets the object use condition. If it is determined according to the noise detection information that there is noise in the to-be-detected image, it indicates that the image output quality of the sensor is poor and does not meet the object use condition. At this time, the object attribute information of the sensor can be obtained, such as sensor identification information, sensor output environment information, etc. The noise detection information and the object attribute information are generated into corresponding inspection information, and are fed back to the inspection personnel, so that the inspection personnel can exclude the abnormal sensor.

[0196] To sum up, the stripe noise detection method provided by the present disclosure determines the to-be-detected image output by the target object and the channel image group corresponding to the to-be-detected image. The to-be-detected image is split according to the channel color information, which facilitates subsequent image detection on different channel images of the to-be-detected image, and improves the image detection efficiency. The target channel image is selected from the plurality of channel images, the image amplitude spectrum of the target channel image is determined, and the column pixel information and the row pixel information are determined according to the image amplitude spectrum. Subsequently, the target channel image is detected for noise based on the column pixel information and the row pixel information according to the preset detection strategy, and the noise detection result of the channel image group is determined. The image noise detection based on the image amplitude spectrum is realized. Subsequently, the column pixel information and the row pixel information are used to complete the noise detection on the target channel image according to the preset detection strategy, which reduces the detection complexity and improves the detection efficiency. Finally, the noise detection result of the channel image group is used to generate the final noise detection information of the to-be-detected image, so that the noise detection information can more comprehensively reflect the image quality of the to-be-detected image, and reliable and effective inspection data is provided for subsequent inspection of the image output quality of the target object using the noise detection information.

[0197] The following describes the stripe noise detection method in combination with the accompanying Figure 17 The stripe noise detection method provided by the present disclosure is taken as an example in the application of sensor quality inspection. Among them, Figure 17 A process flow diagram of a stripe noise detection method according to an embodiment of the present disclosure is shown, which specifically includes the following steps.

[0198] In step S602, in response to receiving a quality inspection instruction for the target object, the inspection environment information is determined, and the target object performs image output according to the inspection environment information to obtain a to-be-detected image.

[0199] In an embodiment of the present disclosure, the quality of the sensor is inspected. At this time, the detection end receives and responds to the quality inspection quality of the sensor, determines that the inspection environment information is a light source-free inspection, and uses the sensor to perform image output in a light source-free inspection environment. The image output by the sensor is used as the to-be-detected image.

[0200] In step S604, multiple channel color information corresponding to the to-be-detected image is determined, and the to-be-detected image is parsed according to the multiple channel color information to obtain a channel image corresponding to each channel color information, and a channel image group corresponding to the to-be-detected image is constructed according to the channel image corresponding to each channel color information.

[0201] In an embodiment of the present disclosure, four kinds of channel color information R, Gr, Gb and B are determined as the multiple channel color information of the to-be-detected image, the to-be-detected image is parsed and split according to the four kinds of channel color information, a channel image corresponding to each kind of channel color information is obtained, and then the four channel images are constructed into a channel image group corresponding to the to-be-detected image.

[0202] In step S606, a target channel image is selected from the channel image group, Fourier transform is performed on the target channel image to obtain frequency domain data of the target channel image, a low frequency component in the frequency domain data is shifted to obtain target frequency domain data, and an image amplitude spectrum corresponding to the target channel image is calculated based on the target frequency domain data.

[0203] In an embodiment of the present disclosure, one channel image is selected from the channel image group as a target channel image, Fourier transform is performed on the target channel image to obtain frequency domain data of the target channel image, a low frequency component in the frequency domain data is shifted to the center position of the image spectrum to obtain target frequency domain data, and then an image amplitude spectrum can be calculated based on the target frequency domain data.

[0204] In an embodiment of the present disclosure, calculating the image amplitude spectrum based on the target frequency domain data includes calculating an initial image amplitude spectrum corresponding to the target channel image according to the target frequency domain data, and determining clipping information of the initial image amplitude spectrum according to image attribute information of the target channel image. The initial image amplitude spectrum is clipped according to the clipping information to obtain the image amplitude spectrum corresponding to the target channel image. In implementation, the initial image amplitude spectrum is calculated according to the target frequency domain data, a clipping range is determined according to the image size of the target channel image, and the initial image amplitude spectrum is clipped according to the clipping range to obtain the image amplitude spectrum.

[0205] In step S608, a row pixel point set and a column pixel point set corresponding to the image amplitude spectrum are determined, wherein the row pixel point set includes a row pixel point group composed of each row of pixel points in the image amplitude spectrum, and the column pixel point set includes a column pixel point group composed of each column of pixel points in the image amplitude spectrum.

[0206] In an embodiment of the present disclosure, the row pixel point set and the column pixel point set are determined according to the image amplitude spectrum, and subsequently the row pixel information and the column pixel information can be determined based on the row pixel point set and the column pixel point set respectively.

[0207] In step S610, a row pixel projection value corresponding to each row pixel point group is calculated, and row pixel information corresponding to the target channel image is generated according to each row pixel projection value. A column pixel projection value corresponding to each column pixel point group is calculated, and column pixel information corresponding to the target channel image is generated according to each column pixel projection value.

[0208] In an embodiment of the present disclosure, when the row pixel information and the column pixel information are determined based on the row pixel point set and the column pixel point set respectively, the row pixel projection value of each row pixel point group can be calculated first, and the row pixel information is generated according to each row pixel projection value. Similarly, the column pixel projection value of each column pixel point group can be calculated first, and the column pixel information is generated according to each column pixel projection value.

[0209] In an embodiment of the present disclosure, the process of calculating the row pixel projection value includes determining a target row pixel point group and a row pixel value of each row pixel point in the target row pixel point group. The row pixel projection value of the target row pixel point group is calculated based on the row pixel value of each row pixel point. The row pixel information corresponding to the target channel image is generated according to each row pixel projection value, including: sorting each row pixel projection value in descending order, and selecting a preset number of row pixel projection values as target row pixel projection values according to the sorting result. The row pixel mean and the row pixel standard deviation value are calculated according to the row pixel projection value of each row pixel point group. The target row pixel projection value, the row pixel mean and the row pixel standard deviation value are taken as the row pixel information of the target channel image.

[0210] In an embodiment of the present disclosure, the process of calculating the column pixel projection value includes determining a target column pixel point group and a column pixel value of each column pixel point in the target column pixel point group. The column pixel projection value of the target column pixel point group is calculated based on the column pixel value. The column pixel information corresponding to the target channel image is generated according to each column pixel projection value, including: sorting each column pixel projection value in descending order, and selecting a preset number of column pixel projection values as target column pixel projection values according to the sorting result. The column pixel mean and the column pixel standard deviation value are calculated according to each column pixel projection value. The target column pixel projection value, the column pixel mean and the column pixel standard deviation value are taken as the column pixel information of the target channel image.

[0211] In step S612, an initial noise detection data template is determined according to a preset detection strategy, and the initial noise detection data template is updated based on the column pixel information and the row pixel information respectively to obtain column noise detection data and row noise detection data.

[0212] In an embodiment of the present disclosure, the initial noise detection data template is determined according to a preset detection strategy, the initial noise detection data template is a noise detection inequality, the column pixel information and the row pixel information are respectively brought into the noise detection inequality, and column noise detection data and row noise detection data are obtained, the column noise detection data is the noise detection inequality after the column pixel information is brought in, and the row noise detection data is the noise detection inequality after the row pixel information is brought in.

[0213] In step S614, noise detection is performed on the target channel image according to the column noise detection data and the row noise detection data, and a noise detection result of the target channel image is obtained.

[0214] In an embodiment of the present disclosure, the target channel image is detected according to the column noise detection data and the row noise detection data, that is, whether noise exists in the current target channel image is determined by using the noise detection inequality, so that the noise detection result is obtained.

[0215] In step S616, an initial noise removal data template is determined according to a preset detection strategy, and the initial noise removal data template is updated based on the column pixel information and the row pixel information respectively, and column noise removal data and row noise removal data are obtained.

[0216] In an embodiment of the present disclosure, when noise exists in the image, noise intensity detection can also be performed, and the detection result with weak noise intensity is removed according to the noise intensity. Therefore, the initial noise removal data template, that is, the noise removal inequality, can be determined according to the preset detection strategy, the column pixel information and the row pixel information are respectively brought into the noise removal inequality, and the column noise removal data and the row noise removal data are obtained, the column noise removal data is the noise removal inequality after the column pixel information is brought in, and the row noise removal data is the noise removal inequality after the row pixel information is brought in.

[0217] In step S618, noise removal is performed on the noise detection result according to the column noise removal data and the row noise removal data, and the removed noise detection result is taken as the noise detection result of the target channel image.

[0218] In an embodiment of the present disclosure, the noise detection result is removed according to the row noise removal data and the column noise removal data, that is, the noise intensity of the image in which noise is detected in the noise detection result is detected by using the noise removal inequality, if the noise intensity is less than a preset intensity threshold, the image in which noise exists in the noise detection result can be removed, and the image in which noise exists is adjusted to be an image in which noise does not exist. Finally, the removed noise detection result is taken as the noise detection result of the final target channel image.

[0219] In step S620, noise detection information of the image to be detected is generated using the noise detection result, wherein the noise detection information is used to verify the image output quality of the target object.

[0220] In an embodiment of the present disclosure, after the noise detection result of each channel image is determined, the noise detection information of the image to be detected can be generated, and the image output quality of the sensor can be detected according to the noise detection information.

[0221] In step S622, if it is determined that the target object does not meet the object use condition based on the noise detection information, the object attribute information of the target object is obtained, the verification information is generated based on the noise detection information and the object attribute information, and the verification information is fed back.

[0222] In an embodiment of the present disclosure, after the noise detection information of the image to be detected output by the sensor is determined according to the above steps, if it is determined that there is noise in the image to be detected based on the noise detection information, it means that the sensor does not meet the use condition and the quality verification is unqualified. The attribute information of the sensor and the noise detection information of this time can be used to generate the verification information corresponding to the sensor, and the verification information is fed back to the corresponding personnel, so that the subsequent quality inspection personnel can exclude the sensor with quality problems and repair in time.

[0223] Based on the same concept, the present disclosure also provides a stripe noise detection device 700.

[0224] It can be understood that the stripe noise detection device 700 provided by the embodiments of the present disclosure comprises a hardware structure and / or a software module corresponding to the execution of each function in order to realize the above functions. In combination with the units and algorithm steps of each example disclosed in the embodiments of the present disclosure, the embodiments of the present disclosure can be realized in the form of hardware or a combination of hardware and computer software. Whether a certain function is realized by hardware or computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to realize the described functions for each specific application, but such implementation should not be considered beyond the scope of the technical solutions of the embodiments of the present disclosure.

[0225] Figure 18 A block diagram of a stripe noise detection device 700 according to an embodiment of the present disclosure is shown. Referring to Figure 18 The device comprises a determination unit 702, a selection unit 704, a detection unit 706 and a generation unit 708.

[0226] The determination unit 702 is configured to determine a target object output image to be detected, and a channel image group corresponding to the image to be detected, wherein the channel color information of each channel image in the channel image group is different.

[0227] The selection unit 704 is configured to select a target channel image from the channel image group, determine an image amplitude spectrum corresponding to the target channel image, and determine column pixel information and row pixel information corresponding to the target channel image according to the image amplitude spectrum, wherein the pixel directions of the column pixel information and the row pixel information are different.

[0228] The detection unit 706 is configured to perform noise detection on the target channel image according to a preset detection strategy based on the column pixel information and the row pixel information, and determine a noise detection result of the target channel image.

[0229] The generation unit 708 is configured to generate noise detection information of the image to be detected by using the noise detection result, wherein the noise detection information is used to verify image output quality of the target object.

[0230] In an embodiment, the selection unit 704 is configured to determine the image amplitude spectrum corresponding to the target channel image by the following method, which comprises:

[0231] performing Fourier transform on the target channel image to obtain frequency domain data of the target channel image.

[0232] determining an initial image amplitude spectrum based on the frequency domain data, and determining a low-frequency component in the initial image amplitude spectrum.

[0233] performing shift processing on the low-frequency component to obtain the image amplitude spectrum corresponding to the target channel image.

[0234] In an embodiment, the selection unit 704 is configured to perform shift processing on the low-frequency component to obtain the image amplitude spectrum of the target channel image by the following method, which comprises:

[0235] performing shift processing on the low-frequency component to obtain an intermediate amplitude spectrum.

[0236] determining cutting information corresponding to the intermediate amplitude spectrum according to image attribute information of the target channel image.

[0237] performing cutting processing on the intermediate amplitude spectrum according to the cutting information to obtain the image amplitude spectrum corresponding to the target channel image.

[0238] In an embodiment, the selection unit 704 is configured to determine the column pixel information and the row pixel information corresponding to the target channel image according to the image amplitude spectrum by the following method, which comprises:

[0239] determining a row pixel point set and a column pixel point set corresponding to the image amplitude spectrum, wherein the row pixel point set comprises row pixel point groups each composed of row pixel points in the image amplitude spectrum, and the column pixel point set comprises column pixel point groups each composed of column pixel points in the image amplitude spectrum.

[0240] The row pixel projection value corresponding to each row pixel point group is calculated, and the row pixel information corresponding to the target channel image is generated according to each row pixel projection value.

[0241] In an embodiment, the selection unit 704 determines the sub-row pixel information corresponding to any one row pixel point group by the following method, including:

[0242] The target row pixel point group and the row pixel value of each row pixel point in the target row pixel point group are determined.

[0243] The row pixel projection value of the target row pixel point group is calculated based on the row pixel value of each row pixel point.

[0244] The row pixel information corresponding to the target channel image is generated according to each row pixel projection value, including:

[0245] Each row pixel projection value is sorted in descending order, and a preset number of row pixel projection values are selected as target row pixel projection values according to the sorting result.

[0246] The row pixel mean and the row pixel standard deviation value are calculated according to the row pixel projection value of each row pixel point group.

[0247] The target row pixel projection value, the row pixel mean and the row pixel standard deviation value are taken as the row pixel information of the target channel image.

[0248] In an embodiment, the selection unit 704 determines the sub-column pixel information corresponding to any one column pixel point group by the following method, including:

[0249] The target column pixel point group and the column pixel value of each column pixel point in the target column pixel point group are determined.

[0250] The column pixel projection value of the target column pixel point group is calculated based on the column pixel value.

[0251] The column pixel information corresponding to the target channel image is generated according to each column pixel projection value, including:

[0252] Each column pixel projection value is sorted in descending order, and a preset number of column pixel projection values are selected as target column pixel projection values according to the sorting result.

[0253] The column pixel mean and the column pixel standard deviation value are calculated according to each column pixel projection value.

[0254] The target column pixel projection value, the column pixel mean and the column pixel standard deviation value are taken as the column pixel information of the target channel image.

[0255] In an embodiment, the detection unit 706 is configured to perform noise detection on the target channel image according to a preset detection strategy based on the column pixel information and the row pixel information, including:

[0256] According to the preset detection strategy, an initial noise detection data template is determined, and the initial noise detection data template is updated based on the column pixel information and the row pixel information respectively to obtain column noise detection data and row noise detection data.

[0257] According to the column noise detection data and the row noise detection data, noise detection is performed on the target channel image to obtain a noise detection result of the target channel image.

[0258] In an embodiment, the detection unit 706 is further configured to determine an initial noise removal data template according to a preset detection strategy, and update the initial noise removal data template based on the column pixel information and the row pixel information respectively to obtain column noise removal data and row noise removal data. According to the column noise removal data and the row noise removal data, noise removal is performed on the noise detection result, and the removed noise detection result is taken as the noise detection result of the target channel image.

[0259] In an embodiment, the determination unit 702 is configured to determine a to-be-detected image output by a target object and a channel image group corresponding to the to-be-detected image, including:

[0260] In response to receiving a quality inspection instruction for the target object, the inspection environment information is determined.

[0261] The to-be-detected image is obtained by image output of the target object according to the inspection environment information.

[0262] A plurality of channel color information corresponding to the to-be-detected image is determined, and the to-be-detected image is parsed according to the plurality of channel color information to obtain a channel image corresponding to each channel color information.

[0263] According to the channel image corresponding to each channel color information, a channel image group corresponding to the to-be-detected image is constructed.

[0264] In an embodiment, the device further comprises a feedback unit configured to acquire object attribute information of the target object in a case where it is determined based on the noise detection information that the target object does not satisfy the object use condition.

[0265] The inspection information is generated based on the noise detection information and the object attribute information and is fed back.

[0266] As to the device in the above embodiments, the specific manners in which various modules perform operations have been described in detail in the embodiments of the method, and will not be described herein in detail.

[0267] Figure 19 A block diagram of an apparatus 800 for stripe noise detection is shown, in accordance with an embodiment of the present disclosure. The apparatus 800 can be provided as a terminal. For example, the apparatus 800 can be a mobile phone, a computer, a digital broadcast terminal, a messaging device, a game console, a tablet device, a medical device, a fitness device, a personal digital assistant, and the like.

[0268] Referring to Figure 19 The apparatus 800 can include one or more of the following components: a processing component 802, a memory component 804, a power component 806, a multimedia component 808, an audio component 810, an input / output (I / O) interface 812, a sensor component 814, and a communication component 816.

[0269] The processing component 802 usually controls overall operations of the apparatus 800, such as operations associated with displaying, making phone calls, data communications, camera operations, and recording operations. The processing component 802 can include one or more processors 820 to execute instructions to complete all or part of steps of the methods described above. In addition, the processing component 802 can include one or more modules to facilitate interaction between the processing component 802 and other components. For example, the processing component 802 can include a multimedia module to facilitate the interaction between the multimedia component 808 and the processing component 802.

[0270] The memory component 804 is configured to store various types of data to support operations of the apparatus 800. Examples of these data include instructions for any applications or methods operating on the apparatus 800, contact data, phonebook data, messages, pictures, videos, and the like. The memory component 804 can be realized by any type of volatile or nonvolatile storage devices or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic or optical disk.

[0271] The power component 806 provides power to various components of the apparatus 800. The power component 806 can include a power supply management system, one or more power sources, and other components associated with generating, managing, and distributing power for the apparatus 800.

[0272] Multimedia component 808 includes a screen that provides an output interface between the device 800 and the user. In some embodiments, the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen may be implemented as a touchscreen to receive input signals from the user. The touch panel includes one or more touch sensors to sense touches, swipes, and gestures on the touch panel. The touch sensors may sense not only the boundaries of the touch or swipe action but also the duration and pressure associated with the touch or swipe operation. In some embodiments, multimedia component 808 includes a front-facing camera and / or a rear-facing camera. When the device 800 is in an operating mode, such as a shooting mode or a video mode, the front-facing camera and / or the rear-facing camera may receive external multimedia data. Each front-facing camera and rear-facing camera may be a fixed optical lens system or have focal length and optical zoom capabilities.

[0273] Audio component 810 is configured to output and / or input audio signals. For example, audio component 810 includes a microphone (MIC) configured to receive external audio signals when device 800 is in an operating mode, such as call mode, recording mode, and voice recognition mode. The received audio signals may be further stored in memory 804 or transmitted via communication component 816. In some embodiments, audio component 810 also includes a speaker for outputting audio signals.

[0274] I / O interface 812 provides an interface between processing component 802 and peripheral interface modules, such as keyboards, click wheels, buttons, etc. These buttons may include, but are not limited to, home buttons, volume buttons, power buttons, and lock buttons.

[0275] Sensor assembly 814 includes one or more sensors for providing status assessments of various aspects of device 800. For example, sensor assembly 814 may detect the on / off state of device 800, the relative positioning of components such as the display and keypad of device 800, changes in the position of device 800 or a component of device 800, the presence or absence of user contact with device 800, the orientation or acceleration / deceleration of device 800, and temperature changes of device 800. Sensor assembly 814 may include a proximity sensor configured to detect the presence of nearby objects without any physical contact. Sensor assembly 814 may also include a light sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, sensor assembly 814 may also include an accelerometer, a gyroscope, a magnetometer, a pressure sensor, or a temperature sensor.

[0276] Communication component 816 is configured to facilitate wired or wireless communication between device 800 and other devices. Device 800 can access wireless networks based on communication standards, such as WiFi, 2G, or 3G, or combinations thereof. In one exemplary embodiment, communication component 816 receives broadcast signals or broadcast-related information from an external broadcast management system via a broadcast channel. In one exemplary embodiment, communication component 816 also includes a near-field communication (NFC) module to facilitate short-range communication. For example, the NFC module may be implemented based on radio frequency identification (RFID) technology, Infrared Data Association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology, and other technologies.

[0277] In an exemplary embodiment, the apparatus 800 may be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the methods described above.

[0278] In an exemplary embodiment, a non-transitory computer-readable storage medium including instructions is also provided, such as a memory 804 including instructions, which can be executed by a processor 820 of the device 800 to perform the above-described method. For example, the non-transitory computer-readable storage medium may be a ROM, random access memory (RAM), CD-ROM, magnetic tape, floppy disk, and optical data storage device, etc.

[0279] Figure 20 This is a block diagram illustrating an apparatus 900 for stripe noise detection according to an embodiment of the present disclosure. For example, apparatus 900 may be provided as a server. (Refer to...) Figure 20 The apparatus 900 includes a processing component 922, which further includes one or more processors, and memory resources represented by memory 932 for storing instructions, such as application programs, that can be executed by the processing component 922. The application programs stored in memory 932 may include one or more modules, each corresponding to a set of instructions. Furthermore, the processing component 922 is configured to execute instructions to perform the methods described above.

[0280] The device 900 may also include a power supply component 926 configured to perform power management of the device 900, a wired or wireless network interface 950 configured to connect the device 900 to a network, and an input / output (I / O) interface 958. The device 900 can operate on an operating system stored in memory 932, such as Windows Server™, MacOSX™, Unix™, Linux™, FreeBSD™, or similar.

[0281] It can be understood that, in the present disclosure, "multiple" refers to two or more, and other quantifiers are similar. The association relationship of "and / or" describing the associated objects means that there can be three relationships, for example, A and / or B can mean that A exists alone, A and B exist together, and B exists alone. The character " / " generally represents that the associated objects before and after it are in an "or" relationship. The singular forms "a", "said" and "the" are also intended to include the plural forms, unless the context clearly indicates otherwise.

[0282] It can be further understood that the terms "first", "second", and the like are used to describe various information, but these information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other, and do not indicate a particular order or importance. In fact, the expressions of "first", "second", and the like can be completely interchangeable. For example, the first information can also be referred to as the second information, and similarly, the second information can also be referred to as the first information without departing from the scope of the present disclosure.

[0283] It can be further understood that the terms "center", "longitudinal", "transverse", "front", "back", "up", "down", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the embodiments and simplifying the description, and do not indicate or imply that the devices or elements referred to must have a particular orientation, be constructed and operated in a particular orientation.

[0284] It can be further understood that, unless otherwise specified, "connection" includes direct connection between the two without other components, and also includes indirect connection between the two with other elements.

[0285] It can be further understood that, although the operations in the embodiments of the present disclosure are described in a specific order in the drawings, it should not be understood as requiring the operations to be performed in the specific order shown or in a serial order, or requiring all the operations to be performed to obtain the desired results. In a particular environment, multitasking and parallel processing can be advantageous.

[0286] Other embodiments of the present disclosure will be readily apparent to those skilled in the art upon considering the specification and practicing the invention disclosed herein. The present disclosure is intended to cover any variations, uses, or adaptive changes of the present disclosure that follow the general principles of the present disclosure and include common knowledge or conventional technical means in the art that are not disclosed in the present disclosure.

[0287] It should be understood that the present disclosure is not limited to the precise construction that has been described above and shown in the accompanying drawings, and that various modifications and changes can be made by those skilled in the art without departing from the scope of the present disclosure. The scope of the present disclosure is limited only by the appended claims.

Claims

1. A method for detecting stripe noise, characterized in that, include: The target object outputs a detection image and a corresponding channel image group, wherein the channel color information of each channel image in the channel image group is different. Select a target channel image from the channel image group, determine the image amplitude spectrum corresponding to the target channel image, and determine the column pixel information and row pixel information corresponding to the target channel image based on the image amplitude spectrum, wherein the pixel directions of the column pixel information and the row pixel information are different; Based on the column pixel information and row pixel information, noise detection is performed on the target channel image according to a preset detection strategy to determine the noise detection result of the target channel image; Using the noise detection results, noise detection information of the image to be detected is generated, wherein the noise detection information is used to verify the image output quality of the target object.

2. The method according to claim 1, characterized in that, Determining the image amplitude spectrum corresponding to the target channel image includes: Perform a Fourier transform on the target channel image to obtain the frequency domain data of the target channel image; The low-frequency components in the frequency domain data are shifted to obtain target frequency domain data, and the image amplitude spectrum corresponding to the target channel image is calculated based on the target frequency domain data.

3. The method according to claim 2, characterized in that, Calculating the image amplitude spectrum corresponding to the target channel image based on the target frequency domain data includes: The initial image amplitude spectrum corresponding to the target channel image is calculated based on the target frequency domain data, and the truncation information of the initial image amplitude spectrum is determined based on the image attribute information of the target channel image. The initial image amplitude spectrum is truncated according to the truncated information to obtain the image amplitude spectrum corresponding to the target channel image.

4. The method according to claim 1, characterized in that, Determining the column pixel information and row pixel information corresponding to the target channel image based on the image amplitude spectrum includes: Determine the row pixel set and column pixel set corresponding to the image amplitude spectrum, wherein the row pixel set includes a row pixel group composed of each row pixel in the image amplitude spectrum, and the column pixel set includes a column pixel group composed of each column pixel in the image amplitude spectrum; Calculate the row pixel projection value corresponding to each row pixel group, and generate the row pixel information corresponding to the target channel image based on each row pixel projection value. Calculate the column pixel projection value corresponding to each column pixel group, and generate the column pixel information corresponding to the target channel image based on each column pixel projection value.

5. The method according to claim 4, characterized in that, The row pixel projection value corresponding to any row pixel group is determined in the following way: Determine the target row pixel group, and the row pixel value of each row pixel in the target row pixel group; Calculate the row pixel projection value of the target row pixel group based on the row pixel value of each row pixel; Generate row pixel information corresponding to the target channel image based on the projection value of each row pixel, including: Sort the row pixel projection values ​​in descending order, and select a preset number of row pixel projection values ​​as the target row pixel projection values ​​based on the sorting results. Calculate the row pixel mean and row pixel standard deviation based on the row pixel projection values ​​of each row pixel group; The target row pixel projection value, the row pixel mean, and the row pixel standard deviation are used as the row pixel information of the target channel image.

6. The method according to claim 4, characterized in that, The column pixel projection value corresponding to any column pixel group is determined in the following way: Determine the target column pixel group, and the column pixel value of each column pixel in the target column pixel group; Calculate the column pixel projection value of the target column pixel group based on the column pixel value; Generate column pixel information corresponding to the target channel image based on the projection value of each column pixel, including: Sort the column pixel projection values ​​in descending order, and select a preset number of column pixel projection values ​​as the target column pixel projection values ​​based on the sorting results. Calculate the column pixel mean and column pixel standard deviation based on the projected values ​​of each column pixel; The target column pixel projection value, the column pixel mean, and the column pixel standard deviation are used as the column pixel information of the target channel image.

7. The method according to claim 1, characterized in that, Based on the column pixel information and row pixel information, noise detection is performed on the target channel image according to a preset detection strategy to determine the noise detection result of the target channel image, including: An initial noise detection data template is determined according to a preset detection strategy, and the initial noise detection data template is updated based on the column pixel information and row pixel information respectively to obtain column noise detection data and row noise detection data. Based on the column noise detection data and the row noise detection data, noise detection is performed on the target channel image to obtain the noise detection result of the target channel image.

8. The method according to claim 7, characterized in that, After generating the noise detection results for the channel image group, the method further includes: An initial noise removal data template is determined according to the preset detection strategy, and the initial noise removal data template is updated based on the column pixel information and row pixel information respectively to obtain column noise removal data and row noise removal data. Based on the column noise removal data and row noise removal data, noise removal is performed on the noise detection results, and the noise detection results after removal are used as the noise detection results of the target channel image.

9. The method according to any one of claims 1-8, characterized in that, Determine the target object's output image to be detected, and the corresponding channel image group of the image to be detected, including: In response to receiving a quality inspection instruction for the target object, determine the inspection environment information; The image to be detected is obtained by outputting an image from the target object based on the inspection environment information; Determine the color information of multiple channels corresponding to the image to be detected, and parse the image to be detected according to the multiple channel color information to obtain the channel image corresponding to each channel color information; Based on the channel image corresponding to the color information of each channel, construct the channel image group corresponding to the image to be detected.

10. The method according to any one of claims 1-8, characterized in that, The method further includes: If, based on the noise detection information, it is determined that the target object does not meet the object usage conditions, the object attribute information of the target object is obtained. Based on the noise detection information and the object attribute information, inspection information is generated and fed back.

11. A stripe noise detection device, characterized in that, include: The determining unit is used to determine the image to be detected output by the target object, and the channel image group corresponding to the image to be detected, wherein the channel color information of each channel image in the channel image group is different; The selection unit is used to select a target channel image from the channel image group, determine the image amplitude spectrum corresponding to the target channel image, and determine the column pixel information and row pixel information corresponding to the target channel image based on the image amplitude spectrum, wherein the pixel orientations of the column pixel information and the row pixel information are different. The detection unit is used to perform noise detection on the target channel image based on the column pixel information and row pixel information according to a preset detection strategy, and determine the noise detection result of the target channel image; The generation unit is used to generate noise detection information of the image to be detected using the noise detection results, wherein the noise detection information is used to check the image output quality of the target object.

12. The apparatus according to claim 11, characterized in that, The selection unit determines the image amplitude spectrum corresponding to the target channel image in the following manner: Perform a Fourier transform on the target channel image to obtain the frequency domain data of the target channel image; The low-frequency components in the frequency domain data are shifted to obtain target frequency domain data, and the image amplitude spectrum corresponding to the target channel image is calculated based on the target frequency domain data.

13. The apparatus according to claim 12, characterized in that, The selection unit calculates the image amplitude spectrum corresponding to the target channel image based on the target frequency domain data in the following manner: The initial image amplitude spectrum corresponding to the target channel image is calculated based on the target frequency domain data, and the truncation information of the initial image amplitude spectrum is determined based on the image attribute information of the target channel image. The initial image amplitude spectrum is truncated according to the truncated information to obtain the image amplitude spectrum corresponding to the target channel image.

14. The apparatus according to claim 11, characterized in that, The selection unit determines the column pixel information and row pixel information corresponding to the target channel image based on the image amplitude spectrum in the following manner: Determine the row pixel set and column pixel set corresponding to the image amplitude spectrum, wherein the row pixel set includes a row pixel group composed of each row pixel in the image amplitude spectrum, and the column pixel set includes a column pixel group composed of each column pixel in the image amplitude spectrum; Calculate the row pixel projection value corresponding to each row pixel group, and generate the row pixel information corresponding to the target channel image based on each row pixel projection value. Calculate the column pixel projection value corresponding to each column pixel group, and generate the column pixel information corresponding to the target channel image based on each column pixel projection value.

15. The apparatus according to claim 14, characterized in that, The selection unit determines the sub-row pixel information corresponding to any row pixel group in the following ways: Determine the target row pixel group, and the row pixel value of each row pixel in the target row pixel group; Calculate the row pixel projection value of the target row pixel group based on the row pixel value of each row pixel; Generate row pixel information corresponding to the target channel image based on the projection value of each row pixel, including: Sort the row pixel projection values ​​in descending order, and select a preset number of row pixel projection values ​​as the target row pixel projection values ​​based on the sorting results. Calculate the row pixel mean and row pixel standard deviation based on the row pixel projection values ​​of each row pixel group; The target row pixel projection value, the row pixel mean, and the row pixel standard deviation are used as the row pixel information of the target channel image.

16. The apparatus according to claim 14, characterized in that, The selection unit determines the sub-column pixel information corresponding to any column of pixels in the following ways: Determine the target column pixel group, and the column pixel value of each column pixel in the target column pixel group; Calculate the column pixel projection value of the target column pixel group based on the column pixel value; Generate column pixel information corresponding to the target channel image based on the projection value of each column pixel, including: Sort the column pixel projection values ​​in descending order, and select a preset number of column pixel projection values ​​as the target column pixel projection values ​​based on the sorting results. Calculate the column pixel mean and column pixel standard deviation based on the projected values ​​of each column pixel; The target column pixel projection value, the column pixel mean, and the column pixel standard deviation are used as the column pixel information of the target channel image.

17. The apparatus according to claim 11, characterized in that, The detection unit performs noise detection on the target channel image based on the column pixel information and row pixel information according to a preset detection strategy, including: An initial noise detection data template is determined according to a preset detection strategy, and the initial noise detection data template is updated based on the column pixel information and row pixel information respectively to obtain column noise detection data and row noise detection data. Based on the column noise detection data and the row noise detection data, noise detection is performed on the target channel image to obtain the noise detection result of the target channel image.

18. The apparatus according to claim 17, characterized in that, The detection unit is further configured to determine an initial noise removal data template according to the preset detection strategy, and update the initial noise removal data template based on the column pixel information and row pixel information respectively to obtain column noise removal data and row noise removal data. Based on the column noise removal data and row noise removal data, noise removal is performed on the noise detection results, and the noise detection results after removal are used as the noise detection results of the target channel image.

19. The apparatus according to any one of claims 11-18, characterized in that, The determining unit determines the image to be detected output by the target object, and the channel image group corresponding to the image to be detected, in the following manner: In response to receiving a quality inspection instruction for the target object, determine the inspection environment information; The image to be detected is obtained by outputting an image from the target object based on the inspection environment information; Determine the color information of multiple channels corresponding to the image to be detected, and parse the image to be detected according to the multiple channel color information to obtain the channel image corresponding to each channel color information; Based on the channel image corresponding to the color information of each channel, construct the channel image group corresponding to the image to be detected.

20. The apparatus according to any one of claims 11-18, characterized in that, The device further includes a feedback unit, used to obtain object attribute information of the target object when it is determined, based on the noise detection information, that the target object does not meet the object usage conditions; Based on the noise detection information and the object attribute information, inspection information is generated and fed back.

21. An electronic device, characterized in that, include: processor: Memory used to store processor-executable instructions; The processor is configured to execute the stripe noise detection method according to any one of claims 1 to 10.

22. A storage medium, characterized in that, The storage medium stores instructions that, when executed by a processor, enable the processor to perform the stripe noise detection method according to any one of claims 1 to 10.

Images