427 results about "Color difference" patented technology

Provided in the present invention is a reconstruction and driving method based on a multi-view motion video of a clothed human body. The method comprises: using a two-layer three-dimensional Gaussian representation to model a human body and corresponding clothes; defining geometric attributes of three-dimensional Gaussian ellipsoids for any face on coarse three-dimensional meshes of the human body and the clothes; at a reconstruction stage, transforming coarse triangular meshes of the clothes and the body from a canonical space to an observation space, and updating Gaussian geometric attributes; and on the basis of a camera pose and parameters of an input video, rasterizing a three-dimensional Gaussian representation of the observation space to an image space, and rendering a picture on the basis of a Gaussian splatting formula. Self-supervised training can be performed on a model by means of calculating the color difference between a rendered picture and a real picture and using same as an optimization objective; when reconstruction is completed and driving is performed, a geometric result of a current frame is used to calculate the position of the coarse triangular mesh of the clothes in the next frame; and the reconstruction-stage method and related parameters are used to render a driving picture in a new posture.

The invention relates to the technical field of image processing, and discloses a fluorescent lifetime microscopic image large-view-field splicing method and system and electronic equipment, and the method comprises the steps: constructing a tissue region mask for a to-be-spliced image, and constructing a vignetting model in a tissue region to complete vignetting correction; counting brightness indexes in an organization area to determine a global brightness reference, and realizing image group brightness unification through global zooming and single image brightness adaptive correction; further realizing alignment of adjacent images through feature point matching, and finishing image block splicing in combination with a minimum color difference suture fusion method to obtain an image band; an overlapping area is extracted from an image belt to generate an effective content mask, relative displacement is estimated by using a phase correlation method after low-frequency suppression, image belt splicing is completed by multiplexing a minimum color difference suture fusion method, a large-view-field spliced image is output, the automation degree and robustness of splicing are improved, the spliced image is geometrically consistent and visually seamless, and the splicing efficiency is improved. And the requirements of medical research on high-resolution and large-field-of-view fluorescence lifetime microscopic images are met.

The invention discloses an LED lamp strip dynamic image quality intelligent optimization method and system based on deep learning driving, and relates to the technical field of LED intelligent control. The LED lamp strip dynamic image quality intelligent optimization method and system based on deep learning driving comprises the steps of S1, preprocessing real-time LED working condition data, memory reference data and historical fluctuation abnormal data; s2, adaptively dividing the LED lamp strip into different independent partitions according to spatial and physical attributes through multi-dimensional feature analysis and clustering; s3, based on a time sequence derivative sensitivity theory, judging thermal drift instability and color temperature abnormity, and entering a sudden change criterion process; s4, executing a sudden change criterion process according to a trend sudden change composite criterion result, and entering an offset compensation process; and S5, based on a criterion compensation theory, executing an offset compensation process. The problems of image quality degradation and color difference accumulation after long-term operation due to the fact that an existing LED lamp strip image quality intelligent optimization system is difficult to achieve real-time self-adaptive compensation of color temperature drifting and color cast phenomena caused by chip junction temperature rise are solved.

The invention discloses a printing color closed-loop control system based on spectral analysis, and relates to the technical field of computer processing, and the system comprises the following steps: S01, collecting the spectral reflectivity data of a plurality of key color blocks on a printed matter measurement and control strip in real time through an online high-speed spectrometer at the frequency of not less than 100 color blocks per second; s02, constructing a color prediction model, wherein the model receives the spectral reflectivity data and outputs absorption scattering characteristics, the spectral transmittance of ink and dot expansion, trapping and overprinting effects in the printing process; s03, setting a dynamic tolerance threshold value delta Et based on a chromatic aberration formula according to the color precision requirement of the printed matter, and judging the absorption and scattering characteristics, the spectral transmittance of the ink and a grading tolerance interval in which the dot expansion, trapping and overprinting effects in the printing process are consistent with visual perception; and S04, converting the spectral reflectivity data acquired in real time into a CIELAB color space irrelevant to equipment, and comparing the spectral reflectivity data with a preset standard color target value. According to the method, the problems of different standards and color drift caused by artificial experience difference and visual fatigue are solved.

The invention relates to the technical field of image data processing, in particular to a high-voltage switch shell surface coating uniformity evaluation method and system, and the method comprises the steps: collecting a shell surface coating image; dividing the image into a plurality of super-pixel areas by using a super-pixel segmentation algorithm introducing an adaptive distance measurement mechanism; extracting geometric structure features, texture features and adjacent color difference features of the target area; and carrying out nonlinear fusion on the adjacent chromatic aberration and geometric structure characteristics to construct inter-class characteristics, fusing the inter-class characteristics with texture characteristics serving as intra-class characteristics to obtain a comprehensive score, and judging whether the coating is uniform or not according to the comprehensive score. According to the method, the segmentation weight is adjusted in a self-adaptive manner, and the weak chromatic aberration is amplified in a nonlinear manner, so that accurate segmentation of a tiny defect region is realized; geometric irregularity, color mutation and texture anomaly features are integrated, detection of various tiny coating defects is achieved, and the evaluation accuracy is improved.

The invention relates to the technical field of computer vision, in particular to a fabric color fastness analysis method and system based on image processing. Comprising the following steps: acquiring a multispectral image of a to-be-detected fabric, and synchronously acquiring multi-factor data; processing the multispectral image to generate an enhanced image; performing image recognition on the gray sample card in the enhanced image to generate a calibration image; performing spectral data projection on the calibration image through a dimension reduction algorithm, and performing color analysis and quantification on color change to generate color difference features; extracting texture features by using a conditional generative adversarial network, carrying out image recognition and calculation on the calibration image through a decoupling algorithm, and generating a region credibility graph; and inputting the chromatic aberration features, the texture features, the multi-factor data and the regional credibility map into an image feature fusion model for mapping, performing analysis through gradient visualization, and outputting an abnormal feature saliency map. According to the method, a color fastness objective analysis closed loop is created, so that the accuracy and the universality of an analysis result are improved.

The invention relates to the technical field of image recognition, in particular to a digital printing chromatic aberration automatic detection and adjustment method based on AI image processing, which comprises the following steps of: calibrating actually measured chromatic aberration and parameter offset to construct a lookup table, collecting an image to be detected to calculate real-time chromatic aberration, and calculating the real-time chromatic aberration; and based on the lookup table, extracting the parameter offset of the minimum distance by using the Manhattan distance to generate an adjustment instruction, sending the instruction to adjust the ink, and collecting new data to update the lookup table. According to the method, the color difference vector and parameter offset mapping lookup table is constructed in the calibration stage, direct association of the color deviation and the hardware adjustment instruction is achieved, the real-time color difference vector is rapidly matched with the Manhattan distance in the detection process, the optimal parameter offset is extracted to generate the adjustment instruction, manual interpretation and manual configuration are replaced, and the detection accuracy is improved. And after execution, a new image is collected to update the lookup table, so that the system continuously optimizes and adjusts the strategy, the detection instantaneity and the closed-loop automation level are improved, and accurate and stable color rendition is ensured.

The invention belongs to the technical field of defect detection, and particularly relates to an artificial quartz stone plate surface defect detection system and method to solve the technical problem that normal textures are easily misjudged as defects in an existing detection method. The detection system comprises an acquisition module which combines color components and anisotropy degrees into multi-dimensional features and arranges the multi-dimensional features of all pixels into an observation matrix; the construction module is used for decomposing the observation matrix into linear superposition of a global background low-rank matrix, a local texture low-rank matrix and a defect sparse matrix; and the extraction module is used for calculating the area of each candidate defect connected domain, the length-width ratio of a bounding rectangle, the average chromatic aberration and the average anisotropy degree. According to the invention, the screening standard can be obtained according to the texture characteristics of the candidate region, pseudo defects caused by noise or background residue can be reduced, and high accuracy and high reliability of the detection result are ensured.

The invention discloses a thermosensitive color-changing modifying pen and a color fixing device. The device comprises a modifying pen and a portable color fixing device. A main cavity and an auxiliary cavity which are separated by a partition plate are arranged in a penholder of the modifying pen and are respectively filled with thermosensitive covering ink and colorless fixation accelerant, and a fine fiber pen point and a metal needle tube pen point which are communicated with the penholder are arranged at two ends of the penholder. The portable color fixing device comprises a constant-temperature heating piece, a power supply unit and a control circuit, according to the principle that thermosensitive covering ink changes color under heating to be matched with the base color of the blueprint, fine smearing and controllable heat fixation technologies are combined, rapid, traceless and lossless modification of the engineering blueprint is achieved, and the problems that a traditional modification method is obvious in color difference, the blueprint is damaged, and operation is tedious are effectively solved. The device is easy and convenient to operate and firm in modification effect, and subsequent use of drawings is not affected.

An image processing apparatus performs a process including acquiring first image data, calculating a white balance value used for an automatic white balance adjustment according to the first image data, performing an automatic white balance adjustment on the first image data according to the white balance value to generate image data after a first correction, calculating a first luminance in the image data after the first correction by adding a color difference component to a luminance, and detecting a first region including a pixel having the first luminance higher than a first threshold value, calculating a correction degree of color information of a pixel included in the first region in the image data after the first correction, according to the first luminance, and correcting the color information of the pixel included in the first region in the image data after the first correction, according to the correction degree.

The application discloses an intelligent color separation method, application method and system, avoids instability caused by human factors, reduces error probability, improves color separation efficiency, accuracy and color uniformity, is simple, easy to learn, high in efficiency and good in accuracy in actual application, can help users improve color control ability, screen high-quality suppliers and improve product quality. The technical scheme is as follows: firstly, taking colors to be separated as standards, corresponding color difference values and related parameters are calculated respectively; secondly, taking the colors to be separated as standards, ranges of related parameters meeting color difference threshold requirements are calculated respectively; and then, colors meeting the related parameter ranges and meeting the color difference threshold requirements are screened out as the same group, and final color separation results are obtained after optimization. The color separation method can be applied to intelligent color separation and evaluation of suppliers of production colors.

The invention relates to the technical field of image processing, in particular to a textile fabric texture area color fastness detection method and device, and the method comprises the steps: collecting RGB images before and after a test during the color fastness detection of the same textile fabric; the two images are grayed and then overlapped parts of edge contours are closed to screen out texture areas of the RGB images; determining brightness features of target pixel points in the RGB image before and after the test, identifying random noise points and constructing range weights of the random noise points, and correcting the brightness features for feature weighting of Gaussian kernel standard deviation during calculation of filtering at the corresponding pixel points; calculating a color difference value of the two filtered RGB images before and after the test in a texture area in the CLELAB color space; and outputting the color fastness grade of the texture area according to the color difference value and the color fastness grading method table. The invention aims to improve the precision of subsequent color fastness detection of the texture area of the textile fabric.

The invention discloses a method and a system for measuring the layer thickness of a multilayer hollow pipe fitting based on image recognition, and relates to the technical field of multilayer hollow pipe fitting detection, and the technical key points are as follows: firstly, enhancing boundary features through multi-band image fusion preprocessing, and offsetting dimensional deviation by combining dynamic environment compensation and residual stress correction; through multi-dimensional verification and error correction, the measurement precision reaches a preset high-precision standard and is obviously superior to that of traditional vernier caliper measurement and ultrasonic measurement; secondly, a stable chromatic aberration boundary is constructed through a color master adaptation rule, and the influence of material pores, defects and distribution uniformity is avoided; meanwhile, through the technologies of self-adaptive noise reduction, dynamic measurement density adjustment and the like, the influence of factors such as ambient light interference and boundary complexity is resisted, and the measurement reliability is ensured; in addition, the color masterbatch adaptation rule can be dynamically adjusted according to optical characteristics and forming process parameters of different materials, the method is suitable for various complex multilayer hollow pipe fittings, and a measurement scheme does not need to be redesigned for specific pipe fittings.

The present application belongs to the technical field of printing, and provides a method, device, equipment and storage medium for eliminating color difference in printing. The method comprises the following steps: obtaining initial printing data obtained after color management processing of a to-be-printed image; obtaining the concentration of each pixel point of the to-be-printed image according to the initial printing data; adjusting the concentration of each pixel point of each pixel row according to the cross-sectional perimeter of the position of the pixel row on a target object; and modifying the initial printing data according to the concentration of each pixel point obtained after the adjustment. The present application further comprises a device, equipment and storage medium for executing the above method. The present application can effectively eliminate color difference in printing.

To perform color mapping to a printing color gamut so as to reduce the degree Iga of color conversion caused by color conversion, and to perform appropriate mapping even when a printing original is a read original.SOLUTION: It is determined whether the first image is an image obtained by scanning a print image. When the first image is an image obtained by scanning a print image, color conversion processing for converting a first color and a second color included in the first image and defined by a first color gamut into a third color and a fourth color defined by a second color gamut different from the first color gamut is executed. When a color difference between the third color and the fourth color is smaller than a predetermined threshold, a conversion parameter in the color conversion processing is corrected so that a color obtained by converting the first color becomes a fifth color different from the third color, in which a color difference from the fourth color is larger than the color difference between the third color and the fourth color.SELECTED DRAWING: Figure 21

The invention discloses an industrial product surface pattern abnormity rapid identification and positioning method and system, and belongs to the technical field of abnormity identification, and the method comprises the steps: collecting industrial product surface image data based on a high-resolution camera in cooperation with a light source device; preprocessing the industrial product surface image data, extracting feature information, and determining industrial product surface image features; an industrial product surface pattern abnormity rapid identification model is constructed to carry out analysis and mode identification on the industrial product surface image features, and an industrial product surface pattern abnormity rapid identification result is determined; and positioning the rapidly identified abnormal area of the surface pattern of the industrial product, and generating accurate position information of the abnormal area of the surface pattern of the industrial product. The problems that existing product quality detection is low in efficiency and high in error rate are solved. The method can quickly and accurately identify and position the abnormity of the industrial product surface pattern, such as scratches, stains and chromatic aberration, so as to improve the product quality detection efficiency and reduce the labor cost and the error rate.

The invention relates to a three-dimensional nondestructive measurement method for Lanzhou lily volume, which comprises the following steps: (1) optimal view selection: constructing Lanzhou lily multi-dimensional evaluation indexes by using information entropy, new point rate, color difference, global features and view angle redundancy, selecting an image with the maximum information entropy, adding an optimal view set, and selecting an optimal view; sending into a three-dimensional reconstruction model to obtain external parameters, internal parameters and a depth map of the camera, and constructing an initial point cloud through coordinate transformation; (2) constructing a comprehensive evaluation function, and selecting three optimal views from the candidate view set in an incremental iteration mode; (3) establishing a three-dimensional reconstruction neural network model comprising a structured coding layer, a dual-channel collaborative reasoning module and a plurality of special interpreters; (4) performing 3D feature extraction and volume prediction: extracting three types of shape features, namely shape parameters, geometric proportion parameters and shape indexes, of the Lanzhou lilium brownii; and then establishing a BPNN neural network model, and predicting the lily volume. The method has the characteristics of rapidness and high precision.

The application discloses an elevator control method, a main control system, electronic equipment and a storage medium, comprising: when the elevator receives goods at an outer call floor and closes the door, a first image under at least two light modes is obtained, a goods area is determined based on the first image, and a current light mode of the car is determined, different light modes are transformed so that the car bottom and side in the first image and the goods have a large color difference, and then the goods area and the corresponding light mode can be determined based on the first image, and the goods area obtained under different light modes may be different. When the elevator arrives at the goods delivery floor and opens the elevator door, a second image of the inside of the car is obtained, whether the goods corresponding to the goods area are delivered is judged based on the goods area and the second image, and after it is determined that the goods are delivered, the elevator door is controlled to be closed and the outer call of other floors is responded, the time taken to close the door is short, unnecessary door opening time can be reduced, and the use efficiency of the elevator is improved.

Method of obtaining a color subpalette from a first moodboard containing one or more digital color images, said method comprising the steps of: (1) generating a first color palette with k1 representative colors from the first moodboard by a method that takes account of underrepresented colors, (2) providing a second color palette with k2 colors, (3) ranking the colors in the first color palette according to a scoring algorithm, which assigns a score to each of the colors in the first color palette based on its color difference with each of the colors in the second color palette, and (4) selecting a subset of ksub colors from the first color palette, which have the scores corresponding to the least color difference with the colors in the second color palette, said subset of colors being the color subpalette. The disclosure also provides a data processing device and a computer program product.

This invention relates to a heatmap evaluation method and system based on class activation mapping and dynamic weights, comprising: acquiring crop images; inputting the crop images into a pre-trained deep learning model to output crop detection results; processing the crop detection results into a heatmap based on class activation mapping, and performing noise reduction processing on the heatmap to obtain a heatmap without background interference; performing color space conversion on the heatmap without background interference, and calculating the color difference between each pixel and a reference color, wherein the reference color includes a preset reference color and corresponding weight based on the heatmap; assigning a weight value to each pixel based on the color difference between each pixel and the reference color, and calculating an average temperature weight value; and completing a quantitative evaluation of the heatmap based on the average temperature weight value. This invention provides a new interpretive dimension and quantitative standard for the application of deep learning models in agriculture.

The present application relates to the field of light box, disclose a kind of visual light box, including box and the shielding portion and the knob being arranged in the inside of the box, the shielding portion and the knob are respectively rotationally connected with the box, part of the shielding portion is arranged on the rotation path of the knob, the shielding portion covers the lighting assembly in the box, by rotating the knob to drive the shielding portion selectively shielding the lighting assembly or display inside the box, the shielding portion when shielding the display, the shielding portion is locked by the knob, by algorithm calculation, the color identification of pattern, and the comparison of color difference between different patterns is carried out, and the color result is output to the display screen;In addition to using camera to cooperate with lighting assembly detection, the present application is also compatible with artificial detection, i.e. through the observation of pattern color by human eye, so that the applicability of the present application is stronger.

The application discloses a kind of tea rolling degree classification methods based on machine vision system, comprising the following steps: step 1: based on machine vision system acquisition different rolling time under tea rolling leaf sample image;Step 2: extract and analyze the color feature information of tea rolling leaf sample image;Step 3: based on portable color difference instrument acquisition different rolling time under tea rolling leaf sample color appearance index;Step 4: based on the color appearance index under different rolling time, obtain the variation law of color appearance index with rolling time;Step 5: based on color feature information and the variation law of color appearance index with rolling time, the classification model of tea rolling degree is established by CNN-SVM algorithm, the application based on machine vision system establishes objective quantization tea rolling degree classification method, realizes the quality quality of product and the quality quality of product of precision control tea processing process.

The invention discloses an RGB-LED spectrum closed-loop system for colorimetry teaching and a color matching method. The system comprises a multi-source input module, a spectrum database module, a dynamic modeling module, an optimization algorithm engine module, a duty ratio calculation module, a PWM driving module and a closed loop verification module. The system is provided with a multi-source input interface, target color coordinates can be obtained, and chromaticity conversion of RGB input is supported; on the basis of the constructed high-precision linear model, solving a normalized strength ratio under constraints by combining an optimization algorithm engine and taking minimization of chromatic aberration as a target; calculating an actual PWM duty ratio according to the actual PWM duty ratio; meanwhile, a spectral irradiation sensor is integrated to collect a mixed color spectrum in real time, actual color coordinates are calculated, and an offset connecting line with an arrow between a target point and an actual point is dynamically drawn in a chromaticity diagram, so that a set of complete visual teaching tool is formed. Compared with a manual color matching scheme, the color mixing precision and the color matching efficiency are greatly improved, and the teaching requirements of colorimetry can be better met.

The present application provides an image processing method, device and storage medium, comprising: acquiring a to-be-tested image of each to-be-tested object in a plurality of same to-be-tested objects, acquiring a definition difference value between a first target region of a first to-be-tested image and a second target region of a second to-be-tested image, acquiring a target parameter according to the definition difference value, and if the target parameter is greater than a target threshold, determining that the first target region and / or the second target region is an image blur region. The present application compares the definition difference value between the target regions at the same position in different to-be-tested images to determine the image blur region, automatically determines the image blur region, improves the identification efficiency of the image blur region, avoids image blur judgment errors caused by background color difference, improves the accuracy of image blur region determination, and the identified image blur region can be subsequently adjusted smoothly to achieve clear detection.

This invention belongs to the field of analytical chemistry technology, specifically relating to a composite indicator and testing method for determining cadmium content in crude cadmium. The composite indicator comprises xylenol orange, thymol blue, and sodium indigo disulfonate, with a mass ratio of xylenol orange, thymol blue, and sodium indigo disulfonate of 1.0–1.5:0.3–0.6:0.5–1.0. This invention utilizes the ternary compound of xylenol orange, thymol blue, and sodium indigo disulfonate, employing the blue background of sodium indigo disulfonate and the reddish-purple and yellow hues of xylenol orange for optical color adjustment. This transforms the color change at the titration endpoint from the traditional reddish-purple to yellow to a bluedish-purple to yellowish-green, significantly increasing the color difference and creating a clear and sharp color change interface, greatly improving the accuracy and reliability of visual endpoint determination.

The invention belongs to the technical field of agricultural product quality nondestructive testing, and relates to a food freshness intelligent monitoring device and method based on olfactory visualization. The device takes an olfaction visualization technology as a core, combines a deep learning algorithm and an Internet of Things technology, and mainly comprises portable detection equipment and an intelligent management cloud platform based on the Internet of Things; wherein the portable detection equipment adopts a main control-edge computing-sensing terminal three-stage heterogeneous framework, and a separated sample reaction chamber is designed, so that real-time acquisition of images of the olfactory sensor is realized. In the monitoring process, firstly, images before and after reaction of a sample and an olfactory sensor are collected, color difference extraction is carried out on the original images, an olfactory sensor data set is established, and a FousNet model is trained by utilizing the data set; and finally, accurately predicting the content of physicochemical indexes of the freshness of the food by utilizing a FousNet model, monitoring and evaluating the freshness of the food in real time, and providing technical support for intelligent management of the freshness of the food.

The present application discloses a method for spectral prediction and closed-loop feedback wide-width corrugated box extended gamut flexographic pre-printing, belonging to the field of packaging engineering and printing technology. The method comprises: establishing a dot gain mapping function based on the Yule-Nielsen correction model by testing the roughness, ink absorption and surface energy of the corrugated paper; printing a CMYKOGV seven-color characteristic color card to generate an adaptive ICC profile; using the CIEDE2000 color difference formula and constant hue angle mapping algorithm for gamut compression and color separation; selecting a low-hardness flat-top dot flexographic plate and configuring a anisotropic screen roller; introducing the Walker-Fetsko rheological ink transfer equation during printing, combining with real-time feedback of an online spectrophotometer to dynamically adjust the printing pressure and ink viscosity, and realizing closed-loop compensation. The present application realizes high-precision simulation of more than 90% of the PANTONE spot color library, completely gets rid of the dependence on spot colors, significantly improves the color stability and production efficiency of wide-width flexographic printing, and is suitable for extended gamut pre-printing production of high-quality corrugated boxes.

This invention discloses an energy-saving color matching method and apparatus for electronic maps based on semantic relationships. The method includes: spatializing the colors of the original map; modeling the map colors as discrete point clouds in the CIELab color space according to perceptual uniformity rules; analyzing the discrete point clouds using a clustering algorithm based on color difference rules to extract semantic relationships between map colors, including differences, associations, and orders; establishing a color semantic relationship evaluation model based on the extracted semantic relationships; reorganizing the colors in the RGB color space and establishing an energy consumption estimation model based on the red, green, and blue component values ​​of each color; establishing the map energy-saving problem as a multi-objective, multi-constraint optimization problem; using a heuristic search algorithm to find energy-saving colors; and obtaining the optimal energy-saving color matching scheme by analyzing the Pareto front solution set. This invention is applicable to vector and raster electronic maps of different scales and themes, automatically extracting semantic relationships to reduce map energy consumption.

The present application relates to the technical field of intelligent detection and automation, and discloses a kind of surface cleaning device's jet flow nozzle fault on-line detection method.Just determine the pixel cleaning area of each nozzle in image;Collect historical images before and after strip cleaning and construct difference map based on color difference, establish surface cleaning quality detection model in combination with the highest and lowest original score, and set qualified standard;Real-time acquisition of images before and after cleaning input model score, identify unqualified images according to qualified standard and mark unqualified area;Semantic segmentation and morphological processing are carried out on the marked area, the contour boundary is extracted, and the faulty nozzle is located in combination with the pixel cleaning area.The method solves the problem of difficult positioning of nozzle fault in existing high-pressure abrasive water jet surface cleaning device.

This invention discloses an online detection system and method for yarn hairiness in texturing machines based on machine vision, belonging to the field of textile automation detection technology. The system includes a pneumatic yarn guiding and cleaning unit, a dual-band color difference imaging unit, and an image processing unit. The pneumatic yarn guiding and cleaning unit utilizes an airflow guiding structure to generate a high-speed laminar flow parallel to the optical window. This high-speed laminar flow simultaneously performs the functions of cleaning the window and constraining the yarn within an air cushion layer using hydrodynamic pressure. The dual-band color difference imaging unit utilizes axial color difference characteristics to focus the first band on the window surface and the second band on the air cushion layer. The image processing unit uses the first band data to quantify the features of deposits on the window surface and performs pixel-level suppression on the second band data accordingly. This invention solves the imaging detection problem caused by the coupling of yarn vibration and stain interference in the high-speed oil mist environment of a texturing machine through the physical coupling of pneumatic constraint and spectral layering.