278 results about "Color detection" patented technology

A recommendation system identifies resale goods corresponding to retail goods being viewed by a user on a retail goods website. Retail goods metadata and images are extracted from the retail goods website using heuristics and LLM-based methods. The extracted retail goods data undergoes ML model-based product category classification, intelligent image cropping, and color detection. Vector embeddings are generated for images and text using ML models and compared to resale goods vector embeddings stored in a vector database. Multiple result sets are retrieved, fused, and re-ranked using LLM-based and preference-aware re-ranking. A data pipeline continuously loads, cleans, and processes resale goods inventory from resale goods websites.

The invention relates to the technical field of image recognition, in particular to an agricultural crop disease and pest detection method and system based on image segmentation, and the method comprises the steps: collecting a vertical view angle image and an inclined view angle image of a tea garden through an unmanned aerial vehicle, carrying out the image registration through SIFT feature extraction, KNN matching and RANSAC filtering, and carrying out the alignment to a reference image coordinate system. And segmenting the foreground region by adopting a U-Net algorithm, and performing post-processing optimization. Positioning a disease area through a color detection method, extracting disease features, and identifying a disease stage by using a CNN algorithm; and using a YOLOv3 algorithm to identify insect body areas, extracting insect pest features, and evaluating insect pest degrees. And finally, based on the disease stage and the pest degree, using an LSTM algorithm to predict the pest diffusion trend, and generating a space thermodynamic diagram according to the reference image coordinate system, thereby improving the detection precision and timeliness, and providing technical support for intelligent agricultural prevention and control.

Embodiments of the present application relate to the technical field of image recognition, and disclose a color band resistor color extraction method and apparatus, a device, and a medium, which solve the technical problem in the prior art of poor precision of color band resistor color detection. The color band resistor color extraction method comprises: acquiring an RGB color image of a target color band resistor; acquiring an LAB color image on the basis of the RGB color image; performing transformation processing on an LAB value corresponding to each color type of the color band resistor to acquire a color extraction threshold; and performing color extraction processing on the LAB color image on the basis of the color extraction threshold. The method of the present application effectively improves the precision of color band resistor color detection.

The invention discloses a display light color detection system, and relates to the technical field of low-brightness display light color detection, and the system comprises a dynamic modulation module which is configured to carry out the high-frequency pulse modulation of a dark-state signal of a detection object, and is used for enabling the dark-state signal to carry a modulation component of a characteristic frequency; the time sequence division acquisition unit comprises a high-speed optical shutter assembly and a detector, and the high-speed optical shutter assembly synchronously switches the on-off state of an optical path according to the characteristic frequency; according to the invention, through high-frequency pulse modulation of dark state signals, time-sequence-divided acquisition of noise and signals and time-frequency domain combined decoupling, traditional hardware limitation is broken through, ultra-low brightness signals and stray light interference are effectively separated, 0.01 cd / m < 2 >-level high-precision detection is realized, the detection efficiency is improved by more than 7 times, and the misjudgment rate is significantly reduced. The problems of missing detection of a novel display technology and the yield of an automatic production line are solved, and an active coding-time sequence calibration-cross-domain cooperation framework has remarkable innovativeness and industrial value.

The invention relates to a color detection method and device. The method comprises the following steps: acquiring a first display area image of a display screen to be detected under a preset display color, and determining a target threshold range corresponding to the preset display color; and determining a coarse detection result according to the target threshold range and the target pixel value of the first display area image in the HSV color space. And determining a fine detection result according to the pixel value of the first display area image in the RGB color space under the condition that the coarse detection result is qualified, determining a color detection result of the display screen to be detected according to the fine detection result, and determining a color detection result of the display screen to be detected according to the coarse detection result under the condition that the coarse detection result is unqualified. Wherein the target threshold range is determined according to a plurality of second display area images of the qualified display screen under a plurality of environment parameters and preset display colors. By adopting the method, the accuracy of color detection can be improved.

The invention relates to the technical field of color detection, in particular to a color difference calculation optimization method and system for archaized linen viscose reactive dyeing. Comprising the following steps: monitoring a fabric motion state in real time, calculating a motion speed and displacement, and determining a spectrum acquisition opportunity and a position compensation parameter; illuminating the surface of the to-be-detected fabric at an acquisition position through a polarized light source, and performing polarized spectrum data acquisition in the normal direction of the surface of the fabric to obtain reflection spectrum information; exciting the surface of the to-be-detected fabric by adopting an ultraviolet light source, collecting fluorescence spectrum data, and identifying different fiber areas; calculating fiber density distribution in the fiber area, and constructing a layered weight system; calculating chromatic aberration of the fiber area based on the reflection spectrum information, extracting chromatic aberration characteristics, and measuring and calculating an antique degree quantification result; setting a color difference threshold value standard according to an antique degree quantification result, and adjusting a color difference threshold value according to an evolution trend; and based on the layered weight system and the chromatic aberration threshold, performing weighted calculation on the chromatic aberration value of the fiber region to obtain a comprehensive chromatic aberration evaluation result.

The invention discloses a red date coloring detection method and system based on an image recognition system, and the method comprises the steps: collecting and preprocessing a red date image sample, and obtaining processed data; performing color space conversion on the processed data; performing feature extraction on the processed data after color space conversion; constructing a red date recognition model based on the extracted features; and completing red date coloring detection by using the red date recognition model. Through the steps of red date image sample collection, preprocessing, color space conversion, feature extraction, red date recognition model construction and the like, automatic detection of the red date color changing process is achieved, the detection efficiency and accuracy are effectively improved, the labor cost is reduced, the objectivity and reliability of red date quality grading are enhanced, and the method is suitable for large-scale popularization and application. And a remarkable promotion effect is achieved on improvement of the automation level and the market competitiveness of the red date processing industry.

The utility model relates to the technical field of chilli fragment color detection, and discloses a color detection device for chilli fragment production, which comprises a detection frame and a production frame, a driving mechanism is fixedly arranged on the inner side of the detection frame, a transparent screening mechanism is fixedly arranged at the output end of the driving mechanism, the driving mechanism comprises a connecting line, and the connecting line is connected with the production frame. An output rotating shaft is fixedly arranged on one side of the connecting line, a cross-shaped driving rod is fixedly arranged on a shaft body of the output rotating shaft, the transparent screening mechanism comprises a transparent cover, L-shaped fixing plates are fixedly arranged on the periphery of the inner side of the transparent cover, and the side, away from the output rotating shaft, of a rod body of the cross-shaped driving rod is connected to the inner side of the transparent cover. The picking mechanism guides a part of crushed chilies in the production process, the crushed chilies are continuously turned over through the transparent screening mechanism so that a chromatographic instrument can conveniently detect the crushed chilies, the sampling time of production personnel and the time for turning over the crushed chilies in the chromatographic instrument are shortened, and the production efficiency is improved.

The invention discloses a textile washing liquor lossless color detection method based on a visible spectrophotometric technique, which comprises the following steps: step 1, pretreatment: washing a textile in a certain manner, taking out the textile from washing liquor after the washing is finished, cooling the washing liquor to normal temperature, and then centrifuging and / or filtering; step 2, testing the transmittance: testing the transmittance of the washing liquid pretreated in the step 1 in a visible light spectrum range; 3, calculating a tristimulus value of the lotion; and step 4, converting the tristimulus value of the washing liquid into related parameters through a uniform color space chromatic aberration formula. According to the color detection method, the operation process is simple and efficient, errors caused by personnel differences, detection parameters and the like are greatly reduced, more objective detection of the washing liquid color is achieved, the detected liquid is not damaged, and the color detection method has very important value and significance for subsequent color fastness rating from the quantization angle.

Disclosed is a surveillance camera capable of detecting a color of a vehicle using vehicle logo detection. This specification describes a method that, instead of processing an entire image of a vehicle, detects the vehicle's logo and expands a bounding box of the detected logo to analyze a dominant color within the expanded bounding box for higher accuracy. During a dominant color analysis process, color correction may be performed based on metadata, environmental factors, and a lighting condition. In addition, more accurate vehicle color detection may be achieved through color distance measurement. According to this specification, at least one of a surveillance camera, an autonomous vehicle, a user terminal, and a server may be associated with an Artificial Intelligence (AI) module, a robot, an Augmented Reality (AR) device, a Virtual Reality (VR) device, and a 5G service.

The invention discloses an intelligent monitoring system for camellia oleifera processing based on artificial intelligence, and relates to the technical field of camellia oleifera fruit processing, and the system comprises a temperature monitoring module which comprises a multi-point temperature sensing array which is arranged along the axial direction of a presser and is installed at the front end, the middle part and an oil outlet of a barrel, the temperature sensor is used for collecting temperature and forming temperature distribution data in the barrel; the color detection module is arranged at an outlet of the oil outlet pipeline, obtains oil chromaticity parameters and a yellow-red ratio through spectral imaging, and extracts color feature data; and the volatile matter detection module is arranged above the discharged oil collection cavity and is used for detecting the concentration of volatile organic compounds through a gas sensing array and generating flavor characteristic data. By arranging the temperature monitoring module, the color detection module and the volatile matter detection module, real-time collection of multi-dimensional information in the squeezing process is achieved, and based on an asynchronous triggering and weighted fusion mechanism, space-time alignment and conjoint analysis are conducted, a joint feature vector is constructed, and a real-time prediction result is output.

The invention relates to an image processing method and device, computer equipment, a storage medium and a computer program product. The method comprises the following steps: acquiring an original image which is not subjected to image signal processing; performing color detection based on the original image to obtain color information corresponding to the original image; and performing image processing on the original image according to the color information. The color gamut where the original image is located does not change, and the original image can well retain the information input to the terminal, so that when color detection is performed based on the original image, the information adjusted by each processing mode of the image signal mode can be used for color detection; therefore, the color information corresponding to the original image can be obtained more accurately for image processing. Besides, in the color detection process of the original image, the color detection can be directly carried out in the color gamut where the original image is located, the color gamut does not need to be converted to the YUV color gamut, the detected color information can be used in the whole image signal processing process, and the problem of frame delay does not exist.

The present invention discloses a method and system for intelligently identifying and locating irregular targets, which are applied to the field of laser marking technology, construct a three-dimensional database, and the three-dimensional database stores three-dimensional data of the target; obtain a three-dimensional image of the target to be identified; extract features from the three-dimensional image to obtain a first positioning feature, and call the positioning feature of the target point; input the first positioning feature into a trained positioning detection model, adjust the position information of the positioning point according to the reflection feature and color feature of the target to be identified, and compare it with the positioning feature; determine the position of the positioning point. The present invention discloses a method and system for intelligently identifying and locating irregular targets, which corrects the acquired reflection feature through the light reflection detection result, clusters similar colors with the color detection result output by the color detection model, determines the laser marking range, and the key point detection result output by the key point detection model jointly locates the key point.

The invention relates to the technical field of detection devices, in particular to a detection device for rapidly measuring the specific gravity, turbidity and color of urine, which comprises a mounting assembly, a through prism, a specific gravity detection unit, a turbidity and color detection unit, a color detection unit and a temperature detection unit, light rays emitted by the specific gravity detection unit enter from one side of the through prism to be totally reflected, receive and collect data, pass through an established fitting function and output a specific gravity value, light rays emitted by the turbidity and color detection unit penetrate through a sample, receive data of scattered light rays passing through the sample, obtain turbidity transmission data and color transmission data, and output a turbidity value in combination with the fitting function. The color detection unit emits light to penetrate through the sample, receives light data reflected by the sample, obtains color reflection data, combines the color reflection data with the color transmission data to establish a fitting function, and outputs a color value, the detection device uses less sample detection, integrates urine specific gravity, turbidity and color detection, is high in automation degree, and improves the detection accuracy.

The invention relates to the technical field of machine vision detection, and discloses a color printed matter quality detection method which comprises the following steps: acquiring original light field response data of the surface of a printed matter in a preset structure light field; carrying out decoupling calculation on the data to generate a basic color graph representing intrinsic colors, a glossiness graph representing process characteristics and a surface normal graph representing geometrical morphology; the method is characterized in that a texture area is identified based on a surface normal graph of a standard sample, and a dynamic reference model is established for the area in combination with a basic color graph so as to provide a self-adaptive color detection threshold value; and finally, detecting color, structure and process flaws in parallel by utilizing each physical characteristic spectrum and the dynamic reference model. According to the method, through decoupling and information guidance of physical dimensions, interference of physical textures on color detection can be fundamentally inhibited, the false alarm rate is remarkably reduced, meanwhile, multi-dimensional comprehensive evaluation of flaws is achieved, and the detection result is objective and reliable.

The embodiment of the invention provides a dust display system and method based on polychromatic light and autonomous cleaning equipment, and belongs to the field of intelligent cleaning equipment. The system comprises a multi-color detection light emitter which is configured to project at least two detection light beams with different wavelengths to a to-be-cleaned ground area in front of or laterally in front of the autonomous cleaning equipment; the light receiver is configured to receive a light signal from the ground area to be cleaned and generate corresponding image data; the control module is electrically connected with the multi-color detection light emitter and the light receiver, and is configured to control the multi-color detection light emitter to switch and emit detection light beams with different wavelengths in a time-sharing mode; and on the basis of multiple groups of image data acquired by the light receiver under the irradiation of the detection light beams with different wavelengths, smudginess identification is performed, and a cleaning strategy control signal is output. The ground background and real dirt can be effectively distinguished, and the false alarm rate is greatly reduced; and meanwhile, the identification capability of smudginess colors and properties is improved, and the method is suitable for diversified family smudginess scenes.

The present invention relates to the technical field of image processing, and in particular to a method and system for analyzing the flame of a hydrogen chloride synthesis furnace, including using a flame sensor to determine whether there is a flame in the hydrogen chloride synthesis furnace; collecting a flame video and obtaining a flame image by frame division; performing color detection on the obtained image under the RGB and HSI mixed color models to obtain a flame pixel point distribution matrix and obtain a preliminary image containing a suspected flame area; performing grayscale processing on the preliminary image containing the suspected flame area, using PSO search for global optimization, and substituting the global optimal solution back into the K-means clustering algorithm to perform clustering on the image; multiplying the matrix of the preliminary image containing the suspected flame area by the matrix of the flame image after clustering to obtain the final flame area; adjusting the gas concentration according to the RGB mean value and color temperature of the final flame area. The present invention solves the problem of poor control quality of manually observing the flame color.

The invention relates to a method for detecting the color of a cell culture solution, and belongs to the technical field of cell detection. The method comprises the following steps: acquiring a color detection value a * actual detection of a to-be-detected cell culture fluid; judging whether the color detection value a * actual detection is smaller than a preset value a * preset of a cell culture fluid or not, and if yes, supplementing or replacing a cell culture medium; otherwise, putting back the culture box to continue culturing for one day, detecting the color detection value of the cell culture solution again on the second day, and supplementing or replacing the cell culture medium until the judgment condition is met. According to the method, the cell growth is not influenced, the color change time of the cell culture fluid can be correctly judged, the cell culture fluid is supplemented or replaced for the cells in time, the detection efficiency is improved, and the labor cost is reduced.

The utility model belongs to the technical field of medical detection, and provides a urine analyzer. The urine color turbidity specific gravity conductivity detection device comprises a sample suction device, a sample conveying pipeline and a urine color turbidity specific gravity conductivity detection device, and the urine color turbidity specific gravity conductivity detection device comprises a shell and a prism module arranged on the shell. The turbidity detection module, the specific gravity detection module, the color detection module and the conductivity detection module are arranged on the prism module, the prism module is provided with a sample channel, and the turbidity detection module, the specific gravity detection module and the color detection module are located in the circumferential direction of the sample channel. The conductivity detection module is positioned at the end part of the sample channel; the two ends of the conveying pipeline are communicated with the sample suction device and the sample channel respectively, so that the sample sucked by the sample suction device is conveyed into the sample channel. The utility model is favorable for reducing the use amount of samples.

The invention relates to an intelligent color box printing color analysis control method, and provides a reinforcement learning decision closed-loop control method integrating distributed color detection, space-time multi-modal data processing, feature extraction and self-supervised clustering and multi-scale reward modeling. Multi-variable normalization and synchronization are completed through multi-point real-time collection of color, process and environment data, color change meta-state features are extracted based on a self-encoding neural network, and a meta-state distribution model related to space and working conditions is established through self-supervised clustering. And the color risk is evaluated in combination with a multi-scale reward function, so that color parameter optimal adaptive adjustment and cooperative control driven by reinforcement learning are realized. In an actual control link, the system can dynamically optimize a meta-state aggregation and strategy network structure based on a feedback signal, a stable and self-adaptive color consistency closed-loop control system is constructed, and the color consistency and process intelligence level of a printed product under a complex working condition are effectively improved.

The invention discloses a traffic light state detection method, apparatus and device, and a computer program product. The traffic light state detection method comprises the steps of obtaining an image of a traffic scene at a current moment; traffic light detection is carried out on the image of the traffic scene at the current moment, a traffic light detection result at the current moment is obtained, and the traffic light detection result at the current moment comprises an initial traffic light color detection result at the current moment; and determining an actual traffic light state at the current moment by using a predefined finite state machine and a preset smoothing mechanism according to the initial traffic light color detection result at the current moment. According to the traffic light state detection method provided by the embodiment of the invention, the traffic light state is identified by using the finite-state machine and the smoothing mechanism of the multi-frame data, which is different from a scheme which only depends on single-frame detection or a single decision model, so that the accuracy and robustness of traffic light state identification are remarkably improved; and identification challenges in a complex traffic environment can be effectively handled.

The present invention relates to a method and a dual-color detection kit for the rapid and simultaneous detection of MS-222 and malachite green, and a technology for the rapid detection of veterinary drug residues in aquatic products. The invention utilizes red and blue colloidal gold with different particle sizes, combined with specific antibodies to construct different red and blue probes, respectively. The invention utilizes the principle that MG-BSA complete antigen and 3-aminobenzoic acid-BSA antigen coated on the T1 line and the T2 line, respectively, can compete with MG and MS-222 in the sample for binding to the blue probe and the red probe, thereby constructing a dual-color rapid detection test strip for MS-222 and malachite green that is clearly distinguishable to the naked eye and simple to operate.

The invention relates to the technical field of hotpot condiment stir-frying, in particular to a caviar hotpot condiment preparation method and equipment thereof.The caviar hotpot condiment preparation method comprises a pot body, a condiment stirring mechanism, a heater, a refrigerator and a quality detection assembly, and the quality detection assembly comprises a supporting plate, a telescopic sampling pipe, a dilution bin and a full-automatic Lovibond colorimeter; a telescopic sampling pipe extends into the pot body, mixed grease of a base material is pumped into a dilution bin, then soybean oil is added into the dilution bin for ten-time dilution, the diluted mixed grease is pumped into a full-automatic Lovibond colorimeter for color detection, and if the color value of Lovibond is within the range of 7.0-8.5, it is indicated that the color is qualified; therefore, manual sampling is not needed, the hotpot condiment can be automatically diluted, the detection efficiency is remarkably improved, color and luster detection is performed through the full-automatic Lovibond colorimeter, the accuracy can be remarkably improved compared with visual inspection, and the color and luster quality of the hotpot condiment is guaranteed.

A system and method is presented for hands-free operation of elevator using a smart phone or like device to display one or more colours that are read by a colour-detection device mounted in an elevator cabin and at each elevator landing. Different colours can be displayed by the smart phone or like device and detected by the colour-detection device where different can represent different commands to operate the elevator.

The invention discloses a fruit quality comprehensive detection method based on machine vision and X-rays. The fruit quality comprehensive detection method is used for performing graded detection on appearance, internal defects and sugar degree of fruits. The appearance detection comprises the steps of constructing a fruit surface defect feature matrix, a fruit shape feature matrix and a color feature matrix, and inputting the fruit surface defect feature matrix into a BPNN network to realize fruit surface defect detection; inputting the fruit shape feature matrix into another BPNN network to realize fruit shape detection; inputting the color feature matrix into an SVM model to realize color detection; the internal defect detection comprises the steps of constructing a plurality of different internal defect detection models, and fusing results of the internal defect detection models based on a multi-criterion decision to obtain an internal defect detection result; the sugar degree detection comprises the steps of extracting a pulp area based on X rays, extracting morphological characteristics and gray statistical characteristics, predicting the sugar degree by utilizing a sugar degree prediction model, and grading the sugar degree according to a sugar degree prediction value. The method can more comprehensively reflect the comprehensive quality of the fruits.

The invention discloses textile color detection equipment which comprises a base, two sets of supporting rods are symmetrically and fixedly arranged on the base, side plates are fixedly arranged on the two sets of supporting rods, two rotating rollers are rotationally arranged between the two side plates, and the two rotating rollers are symmetrically and rotationally arranged at the two ends of the two side plates through shaft rods. The two rotating rollers are sleeved with a conveying belt, a plurality of sets of clamping plates are evenly arranged on the conveying belt, each set of clamping plates comprises two clamping plates symmetrically arranged on the two sides of the conveying belt, each locking bolt is rotationally arranged on the conveying belt, and each clamping plate is provided with a threaded through hole corresponding to the corresponding locking bolt. And a detection chamber is fixedly arranged on the two side plates. The textile can be smoothed before being detected, the situation that the color detection accuracy is affected by shadow generated when the textile is folded is avoided, the detection lens can be quickly wiped when needed, and the color detection quality is guaranteed.

The invention provides a portrait picture processing method and apparatus, and an electronic device. The method comprises the steps of obtaining an uploaded portrait picture; performing clothing feature detection and clothing color detection on the portrait picture to determine a first clothing feature and a clothing color; performing feature matching from a preset clothing asset library according to the first clothing feature, and determining a target clothing asset; and according to the garment color, performing graded dyeing on the target garment assets to obtain virtual role garments. According to the method, the technical capabilities of game clothing asset retrieval and clothing asset grading dyeing can be introduced into an intelligent image creation function, and the problem that in an existing intelligent image creation function, a pinched virtual character image is different from an overall image of an uploaded portrait picture is solved.

The invention discloses an intelligent laser system with an energy-saving function. The intelligent laser system comprises a color detection device, a height measurement device, a laser generation device, a temperature detection device, a conveying device and a controller. According to the device, edge sealing belts with different colors can be detected through the color detection device, edge sealing belts with different heights are measured through the measuring device, and meanwhile, the controller controls the output power of the laser generating device and the size of light spots in real time according to the conveying speed of the conveying device; when the laser generating device outputs laser to process the edge sealing band, the controller detects the temperature of the laser irradiation area through the temperature detection device and adjusts the output power and the light spot size of the laser generating device to form closed-loop control, so that the edge sealing bands with different colors and heights can be processed by proper output power and light spots, and the processing efficiency of the edge sealing band is improved. And a proper processing temperature is achieved, so that higher processing quality is obtained.

The invention discloses a display screen detection system and a display screen detection method, relates to the technical field of display screen detection, and aims to solve the problem of display screen color detection.The display screen detection system comprises a control module, the control module is electrically connected with a processing module, and the processing module is electrically connected with a driving module; the driving module is electrically connected with an adjusting mechanism, and the adjusting mechanism is provided with an industrial camera. The processing module is electrically connected with a calculation model, and the calculation model is used for calculating the collected display screen images, including pixel center distance relative deviation, horizontal relative dislocation and vertical relative dislocation. According to the display screen detection system and the display screen detection method provided by the invention, the processing module is used for preprocessing, and the calculation model is used for calculating the pixel center distance relative deviation, the horizontal relative dislocation and the vertical relative dislocation; and the display screens are sorted through the PID control module.

The invention relates to the technical field of image processing, and discloses a vision-based bad part detection method and device. The method comprises the steps of collecting a target image of a region of interest, wherein the region of interest comprises a standard region where a to-be-detected part is located; detecting the target image based on a part detection model, and determining whether the to-be-detected part belongs to a bad part or not; wherein the part detection model comprises at least one of a part missing detection model, a part offset detection model and a part defect detection model; the part missing detection model detects whether a to-be-detected part is missing or not based on the color of the to-be-detected part, and the part offset detection model is used for detecting whether the to-be-detected part is offset or not based on the part characteristics of the to-be-detected part. The part defect detection model is used for detecting whether the part to be detected has defects or not based on a deep learning model. The detection efficiency, the detection precision and the detection time of defective parts can be improved.