1201 results about "High contrast" patented technology

The invention relates to the technical field of defect detection, and discloses an integrated circuit mask plate defect multispectral cooperative detection method, which comprises the following steps: carrying out preliminary scanning on a mask plate by using a multispectral camera, selecting an optimal wave band combination, switching to a target wave band through a liquid crystal adjustable optical filter, and obtaining a high-confidence multispectral image; the optical path offset is calculated through the linear relation, the deflection angle of the micro-mirror is dynamically adjusted, and images are collected for the second time after imaging offset is compensated; image filtering and registration are completed in the FPGA, and a high-contrast fusion image is generated; if the confidence coefficient is insufficient, triggering a wave band reselection mechanism to recheck; dynamically optimizing a defect type-spectrum mapping table based on real-time detection data; when the environment fluctuation exceeds the limit, an anti-interference mode is automatically switched, and micro-mirror calibration is triggered; and monitoring performance degradation of the optical assembly in real time, switching redundant optical paths, cleaning damaged parts, triggering manual reinspection when conflicts occur, and finally generating a visual report. According to the invention, the detection accuracy of the integrated circuit mask plate can be improved.

The invention discloses a casting detection method and system based on machine vision, and relates to the technical field of intersection of machine vision and material science, and the method comprises the steps: extracting the boundary coordinates of a region with a risk value exceeding a risk threshold according to a defect risk weight map, and forming a dynamic region of interest scanning frame; adjusting the incident angle of a polarized light source according to the track of the dynamic attention area scanning frame, collecting a local image, and obtaining a high-contrast image sequence for eliminating the interference of a surface oxide layer; extracting a defect edge feature map from the high-contrast image sequence through a deformable convolution kernel, and obtaining a defect candidate region mask; the defect candidate region mask is mapped to a von Mises stress field, and a crack germination probability is predicted in combination with a material fatigue limit parameter. According to the method, the accuracy and reliability of detection are improved, and the applicability under complex working conditions is remarkably improved, so that the casting quality control level is effectively improved, and the repair and quality control cost is reduced.

The invention relates to the technical field of machine vision and perception, in particular to a ton bag lifting unmanned control system based on a binocular vision camera and a laser radar, which comprises an intelligent control unit, a lifting appliance executing mechanism, a sensing unit and a special ton bag, the sensing unit comprises a binocular vision camera and a laser radar and is used for collecting depth vision and three-dimensional point cloud information of an operation area; the intelligent control unit fuses multi-source data, locates a lifting lug by improving a weighted multi-feature fusion algorithm, plans a safety path and generates a staged instruction; the lifting appliance executing mechanism lifts and pulls a collapsed lifting lug through an electromagnetic adsorption module, a mechanical gripper module clamps the lifting lug, and reliable operation is achieved in cooperation with a verification mechanism; the special ton bag is matched with a sensing and executing module through a high-contrast color and a pre-embedded metal piece. The full-process unmanned operation is achieved, the robustness and operation safety of the complex environment are improved, and the ton bag hoisting requirements of multiple industries are met.

The invention discloses a multispectral image fusion model based on a double-branch self-attention-generative adversarial network and a fusion method thereof, and the model sequentially comprises an input preprocessing module which is used for carrying out the same-amplitude mapping, normalization and overlapping block embedding of visible light and infrared original images, and generating a to-be-fused feature block; the double-branch encoder module captures a cross-modal long-distance dependence and overall brightness structure through multi-head deep convolution transpose attention (MDTA) and a gated deep convolution feedforward network (GDFN), and extracts high-frequency texture and edge information by using a reversible residual gating layer and detail DetailNode iteration; the fusion decoder module is used for carrying out multi-level self-attention-convolution reconstruction on the two paths of features after channel dimension splicing, and outputting a single-frame high-resolution fusion image; and the double-domain discriminator module comprises a visible light domain discriminator and an infrared domain discriminator which are respectively used for carrying out adversarial evaluation on the fused image and the corresponding modal truth value image so as to improve the detail authenticity and the thermal target contrast ratio of the fusion result. The technical problems that an existing infrared-visible light image fusion method is insufficient in detail reservation, unbalanced in brightness and contrast, poor in unsupervised training stability and the like are solved, the method can be deployed on embedded platforms needing real-time and multi-modal information enhancement such as night monitoring, unmanned driving and edge security and protection, and high-contrast and high-information-amount fusion imaging is achieved.

The invention discloses a coal mine belt deviation real-time lightweight detection device and a coal mine belt deviation real-time lightweight detection method based on image sub-pixels, firstly, the device captures video streams of belt operation in real time through an image acquisition device, analyzes the video streams in real time by means of edge computing nodes, and monitors the running state of a belt in real time to determine the deviation of the belt; once the belt starts to work, image distortion caused by a shooting angle or environment is immediately corrected through perspective transformation. Secondly, automatically selecting a high-contrast reference strip through an image preprocessing method, and accurately extracting a region near the belt edge as a minimum region of interest, namely a belt edge region; then, a lightweight sub-pixel convolutional neural network is used to carry out deep analysis on the areas, and a high-quality super-resolution image is generated, so that the accuracy of edge detection is further improved. And finally, accurately identifying and positioning the edge position of the belt by introducing a bidirectional edge tracking method with an abnormal detection function, thereby realizing real-time and accurate detection on the deviation condition of the belt.

The invention belongs to the technical field of novel fluorescent materials, and particularly discloses application of a mechanical color change pyrenyl derivative in latent fingerprint fluorescence imaging and extraction. The mechanical color-changing pyrenyl derivative can be selectively combined with fingerprint residual substances under the excitation of ultraviolet light through the unique fluorescence response characteristic, so that potential fingerprints form high-contrast fluorescence imaging on smooth or rough surfaces (such as glass, metal, plastic and the like), and the mechanical color-changing pyrenyl derivative is compatible with conventional camera equipment to realize rapid acquisition. The compound design is based on a modular synthesis route, and has the advantages of simple and convenient synthesis process, controllable raw material cost, high reaction yield and the like. The product has excellent thermal stability and chemical stability, can be stored at normal temperature and does not need special treatment, and the practical application threshold is greatly reduced. Compared with a traditional fingerprint developing technology, the method has remarkable breakthroughs in the aspects of sensitivity, substrate adaptability and operation convenience, and an efficient, economical and environment-friendly novel solution is provided for the fields of criminal investigation, forensic identification and the like.

According to the device and the method for synchronously locking the phase and the period of the holographic interference fringes, two sets of locking devices are adopted and are respectively arranged at two ends of a holographic interference exposure field in the same direction as the period of the interference fringes, one set is used for compensating the phase of the interference light field to lock the phase of the interference fringes, and the other set is used for locking the period of the interference fringes. And the other one is used for compensating the output laser frequency of the laser to lock the interference fringe period. According to the interference fringe phase and period synchronous locking method, the problem that the contrast ratio of interference fringes is degraded due to phase and period drift of the interference fringes in the exposure process of a large-aperture holographic grating is solved, and particularly an interference fringe period real-time locking module is used; the problem that the fringe contrast is gradually deteriorated in the periodic direction due to periodic change of interference fringes is effectively solved, and a reliable solution is provided for preparation of the meter-scale large-aperture, high-contrast and high-uniformity holographic grating.

The invention relates to a gray dye composition, which consists of at least one yellow dichroic dye, at least one red dichroic dye, at least one purple dichroic dye and at least one blue dichroic dye, and has uniform total absorption in a visible light region when being mixed with a parent liquid crystal composition for use. The liquid crystal composition has high contrast when applied to light valves, switchable windows, switchable mirrors and guest-host displays.

The invention discloses a beer sealing cover defect identification method based on image intelligent identification, and relates to the field of image identification, and the method comprises the steps: extracting the brightness distribution, color histogram and edge gradient features of a picture shot by a camera; constructing a precise light type calculation model, and dynamically adjusting brightness transition curves of adjacent partitions; identifying a high-contrast edge in the picture, and dynamically adjusting a brightness decay curve of adjacent partitions according to an edge brightness difference; color calibration is carried out through a three-dimensional color lookup table; and performing defect identification on an actual sealing cover image by using a dual mechanism. The light rays of the light source emitter are regulated and controlled, so that under the contrast condition, a machine and human eyes are easy to identify, the halo diffusion radius is reduced, and color calibration is carried out through the three-dimensional color lookup table, so that the machine and the human eyes are prevented from being influenced by images during identification, and the overall identification accuracy is ensured.

The invention discloses a photoelectric wafer defect detection method, and belongs to the technical field of semiconductors. The detection method comprises the steps that the surface of a photoelectric wafer is covered with a quantum dot precursor solution, annealing is carried out, after a quantum dot film is formed, irradiation is carried out under ultraviolet light, a displayed fluorescence signal can mark defects on the surface of the photoelectric wafer, and the purpose of detecting the defects of the photoelectric wafer is achieved. The quantum dots used in the invention have excellent fluorescence efficiency and stability, and can provide a high-contrast marking effect in an SEM image. Therefore, the detection of micron-scale and nano-scale defects is more accurate, and the details are clearer. A traditional fluorescence labeling solution is limited in infiltration capacity in nanoscale defects and can only cover the surface. The quantum dots can penetrate into the defect to form a mark with the depth ratio as high as 1: 3. The depth marking capability is helpful for comprehensively knowing the internal structure of the defect and the defect extension condition. In addition, the quantum dots have excellent adaptability and multifunctionality, and show good stability and coupling effect under different types of wafer conditions.

The invention provides a facial skin flaw enhancement method based on a Lab color space. The method comprises the following steps: firstly, acquiring an RGB face image and converting the RGB face image into a CIE Lab color space with uniform perception; then, performing differentiation treatment according to the manually selected skin flaw type: for the vascular flaw, extracting statistical characteristics of a component and driving adaptive nonlinear transformation, and generating a grey-scale map which highlights the red flaw; for pigment flaws, nonlinear transformation is carried out on the component L, and then collaborative linear weighting and feature amplification are carried out on the component L, the component a and the component b, so that a grey-scale map with highlighted pigment spots is generated. And finally, coloring the grey-scale map in the Lab color space through adjustable parameters to generate a high-contrast color enhanced image. The method overcomes the dependence on hardware and training data in the prior art, can clearly and adaptively enhance various flaws such as acnes, couperose streaks and color spots, shows robustness under different illumination, and can be widely applied to clinical beauty, later photography and real-time video processing.

The resist composition exhibits a high contrast, improved maximum resolution, and satisfactory sensitivity and line width roughness (LWR) when processed by lithography using high-energy radiation, especially electron beam (EB) and EUV, and a pattern forming process using the resist composition. The resist composition comprising (A) a base polymer containing a polymer comprising repeat units having an acid labile group and repeat units having a carboxy group, (B) an organic solvent, and (C) a hypervalent iodine compound having formula (1):

The invention discloses an infrared and visible light image fusion network, and relates to the field of image fusion. The system comprises a double-branch feature coding unit, a fusion unit and an image reconstruction unit, the double-branch feature coding unit comprises a multi-scale edge gradient module, a significant dense residual module and a spatial bias module, and infrared and visible light images pass through the multi-scale edge gradient module to obtain enhanced texture features. The infrared image is processed by a significant dense residual module to obtain enhanced significant features, global information and local information of the enhanced texture features and the strong significant features are fused by a spatial bias module, the features processed by a fusion unit are input into an image reconstruction unit, the fused feature information is reconstructed, and the reconstructed feature information is output to the image reconstruction unit. And outputting a final fused image. According to the method, the clarity of image edges and target details can be better maintained, and higher contrast and visual clarity are provided.

A scan driving method for rendering a high-contrast image, for use in a cholesteric liquid-crystal display device, is provided. The cholesteric liquid-crystal display device includes a display panel and a driving circuit section. The method includes: utilizing the driving circuit section to sequentially activate each scanning electrode using a modified pulse-width modulation (PWM) scanning procedure including a first stage, a second stage, and a third stage in sequence; and utilizing the driving circuit section to apply one or more alternating-current (AC) voltage pulses to pixel circuits on a first scanning electrode during the first stage to manipulate grayscale values of the pixel circuits thereon to be written in the second stage. During the first stage of the first scanning electrode, the pixel circuits on the remaining scanning electrodes within the third stage sense a zero voltage during a high-voltage period within the first stage of the first scanning electrode.

A liquid crystal optical element includes an optical function layer and two transparent substrates. The optical function layer is sandwiched between two transparent substrates. The optical function layer contains liquid crystal molecules and a photoisomerization material. The liquid crystal optical element switches between a light-transmitting state and a light-scattering state according to an orientation state of the liquid crystal molecules. The optical function layer holds the liquid crystal molecules and the photoisomerization material within a mesh structure formed of a resin-particle aggregate. Therefore, light can be strongly scattered when the liquid crystal optical element is in a screen state, and a high-contrast image is obtained when visible light is projected onto the liquid crystal optical element.

The invention relates to the technical field of surveying and mapping, and discloses a surveying and mapping system and method based on multi-source data fusion, and the method comprises the steps: collecting the image data of a to-be-surveyed area, collecting the illumination intensity of the to-be-surveyed area and the inertial data of an unmanned plane in real time, and judging whether to start LiDAR ranging; when it is judged that starting is conducted, image data and point cloud data are obtained, and the point cloud data are processed according to the timestamp of the image data; processing the image data to determine a high-contrast area, and converting image coordinates in the image data into global coordinates; denoising the point cloud data after time synchronization to generate a digital elevation model; and aligning the high-contrast area and fusing other areas in the image data according to the digital elevation model to obtain a surveying and mapping model. By fusing the optical image and the LiDAR data, the limitation of a single surveying and mapping means is overcome, and the precision and reliability of the surveying and mapping model are improved.

The invention provides an ultrasonic transmission CT (Computed Tomography) imaging method and system, a computer and a storage medium, which can be used for quickly reconstructing an ultrasonic transmission image with high resolution and high contrast. The method comprises the following steps: in layer-by-layer scanning of an annular ultrasonic transducer array, calculating a TOF (Time of Flight) matrix containing transmission time of all ultrasonic paths after the annular ultrasonic transducer array rotates by a specific angle each time; after one layer of scanning is completed, the TOF matrixes of different angles are recombined into a three-dimensional matrix according to the spatial relation, and missing data are complemented; discretizing an area to be imaged into two-dimensional grids, making each grid correspond to a pixel point, counting transmission time and path length of all ultrasonic paths penetrating through the pixel point, calculating sound velocity contribution values of the paths, taking an average value as a sound velocity value of the pixel point, and generating a two-dimensional ultrasonic transmission image; a two-dimensional image is converted into a three-dimensional image through a ray casting algorithm, the visual influence of voxels is adjusted by introducing a weight factor based on depth, and the depth perception and semitransparent rendering effect of the image are enhanced.

The invention discloses a red xanthene dye for blue photoresist color paste and a synthesis method of the red xanthene dye, and the synthesis method comprises the following steps: (1) sulfonyl chlorination reaction: carrying out sulfonyl chlorination reaction on xanthene molecules B containing two sulfonate radicals by using a chlorination reagent, and carrying out post-treatment to obtain a sulfonyl chloride intermediate C; and (2) esterification reaction: carrying out esterification reaction on the sulfonyl chloride intermediate C obtained in the first step by utilizing an esterification reagent, and carrying out post-treatment to obtain the target red xanthene dye A with methacrylamide at the tail end. The invention provides a novel xanthene dye molecule which can be mixed with a blue pigment color paste to prepare a blue photoresist color paste for improving the brightness of the color paste. In addition, the liquid crystal composition has the advantages of high transparency, high contrast ratio, good dispersity, relatively high covering power, tinting strength and the like, and is good in thermal stability and light stability; meanwhile, the synthesis method provided by the invention is simple, low in cost and easy for large-scale production.

The invention belongs to the technical field of deep learning and image processing, and discloses an underwater image enhancement method based on tensor learnable prior. By designing learnable prior based on a Tensor Train kernel, explicit modeling of the underwater degradation law is realized. The four Tensor Train kernels correspond to the height, the width, the channel and the image block modality respectively, so that the system can simultaneously describe the vertical structure change, the horizontal scattering, the spectral absorption difference and the regional scale attenuation, thereby obviously enhancing the description capability for the underwater multimode degradation. The Tensor Train kernel is dynamically generated based on the kernel prediction sub-network, the prior structure of the method can be adaptively adjusted along with the input image, and the stability and generalization ability of the enhancement result are improved. According to the method, color shift is effectively corrected and scattering blur is inhibited while the structure edge is kept, so that the enhancement process is changed from traditional black box type mapping to physical consistency structured reasoning, and more natural colors, higher contrast and better detail recovery effects are obtained in a complex turbid environment.

The invention relates to the technical field of machine vision detection, particularly discloses a mechanical part surface defect vision detection system, and aims to solve the problems of blind areas, environment light sensitivity, insufficient tiny defect detection capability and the like in the detection of complex curved surface high-reflection parts. Comprising an optical imaging module, a geometrical morphology sensing module, a dynamic polarization regulation and control module, an image fusion and enhancement module and a defect identification and decision module. The three-dimensional morphology of the surface of the part is sensed in real time, the imaging polarization state is dynamically regulated and controlled according to the three-dimensional morphology, multi-frame polarization image fusion and digital diffraction enhancement processing are combined, full-coverage and high-contrast detection of the complex curved surface part is achieved, motion blur is effectively restrained, and the detection capacity of sub-pixel-level tiny defects and the system environment adaptability are improved.

Provided are a chiral photosensitive molecular switch, a cholesteric liquid crystal composition, and a display panel and a preparation method therefor. The chiral photosensitive molecular switch has two stable states, and can be transformed from one stable state to the other stable state in response to light, and exist stably. The cholesteric liquid crystal composition comprises: the chiral photosensitive molecular switch, a bent-core molecule mixture, a rod-shaped single crystal mixture and a non-photosensitive chiral dopant; the cholesteric liquid crystal composition can be used to prepare a cholesteric liquid crystal display panel, the cholesteric liquid crystal comprising two stable states having different rotation directions, and being able to reflect light having different rotation directions. By adopting the strategy of light regulating the rotation direction of a cholesteric liquid crystal and an electric field regulating a reflection band, it is possible to obtain a cholesteric liquid crystal display panel having high reflectivity, high contrast and good stability, and which can be applied to electric field regulation of reflection in the full spectrum range.

The invention relates to the technical field of computer vision, and discloses a Monte Carlo tree search-based camera lens assembly focusing calibration algorithm and system, and the algorithm comprises the steps: constructing a three-dimensional search space comprising a focus position, a horizontal inclination angle and a vertical inclination angle, and combining a high-contrast cross wire test pattern to generate a local definition matrix; calculating an initial inclination angle based on the matrix to construct a root node of the Monte Carlo tree; according to the algorithm, an optimal search path is dynamically selected by applying an upper confidence bound strategy, child nodes are expanded in a three-dimensional space, physical adjustment is executed, and optimal three-dimensional configuration parameters are finally determined based on node scores through back propagation iterative optimization. The method breaks through a traditional one-dimensional search framework, effectively avoids a pseudo peak trap caused by inclination of a lens-sensor axis, improves the calibration precision and efficiency of the full-view-field definition of the camera, and achieves the precise positioning of a global optimal solution under a complex assembly tolerance.

The invention discloses a vibration table test sample real-time deformation monitoring system based on machine vision. The vibration table test sample real-time deformation monitoring system comprises an active modulation light source module, a multi-camera high-speed acquisition subsystem, a real-time deformation analysis unit, a full-field strain field reconstruction module, a motion compensation fusion module and a vibration synchronous control interface. An active modulation light source is synchronous with vibration, and motion blur is effectively eliminated in cooperation with a narrow pulse width design; the multi-camera orthogonal layout is combined with the vibration isolation base, so that stable acquisition of images is guaranteed; and the high-contrast coding datum point and dynamic tracking module accurately obtains the 6-DOF displacement. The rigid body and the elastic deformation can be accurately separated through a digital image correlation method and motion compensation fusion. High-frequency updating of a full-resolution strain field is achieved through multi-scale feature fusion, and monitoring intuition and safety are improved through an online visualization and early warning module. The system can realize high-precision, real-time and full-field deformation monitoring, meets complex vibration test requirements, and provides reliable data support for material and structure performance evaluation.

The invention discloses an SAR (Synthetic Aperture Radar) image MTF (Modulation Transfer Function) quality evaluation method and system based on a natural scene edge, and the method comprises the steps: building a selection standard system for automatic edge screening of a natural scene SAR image, and determining a high-contrast linear edge according to the standard system; automatically detecting a high-contrast linear edge in the natural scene, and intercepting a region of interest for subsequent calculation based on the position of the linear edge; gray level distribution is extracted in the normal direction of the linear edge, and an edge spread function is constructed and used for describing gray level transition response of the imaging system in the edge area; carrying out normalization and smoothing processing on the edge spread function, and carrying out spatial domain first-order differential calculation to obtain a line spread function; and performing fast Fourier transform on the line spread function to obtain a modulation degree transfer function curve, and representing the detail transfer capability of the system under different spatial frequencies. According to the method, the imaging quality of the SAR image of the natural scene can be automatically, accurately and quantitatively evaluated under the condition that manual field arrangement is not needed.

The invention discloses a focused ultrasound transcranial nerve regulation and control platform and method based on photoacoustic imaging. The system comprises a synchronous pulse generator, a pulse ultrasound transmitting system, a photoacoustic signal collecting system and a fixing device. According to the method, the limitation of other imaging monitoring modes is overcome, monitoring is carried out by utilizing the characteristics of non-invasiveness, high resolution and high contrast of photoacoustic imaging, the TUS progress can be monitored for providing real-time dynamic imaging, and functional information can be provided; photoacoustic imaging is applied to focused ultrasound transcranial nerve regulation for the first time. And meanwhile, by utilizing high sound resolution and high optical contrast of photoacoustic imaging, the strongest photoacoustic signal can be provided in a near-infrared region for hemoglobin of a seat endogenous contrast agent, so that the PAT becomes an ideal imaging tool for detecting cerebral hemodynamics. In addition, key pathophysiological parameters are evaluated through functional imaging provided by the PAT, and the method can be used for evaluating brain changes in the low-intensity focused ultrasound stimulation process.

The present invention belongs to the technical field of mechanochromic materials, and specifically relates to a mechanochromic material containing rhodamine with a hydantoin structure, and its preparation method and application. The rhodamine mechanochromic material containing a hydantoin structure is shown in structure (I), with rhodamine B as the matrix, reduced to rhodamine alcohol, oxidized to generate rhodamine aldehyde, and reacted with hydantoin to prepare a six-membered spirocyclic rhodamine containing a hydantoin structure. The rhodamine compound containing a hydantoin structure involved in the present invention has good ion interference resistance, exhibits photochromic properties in a dichloromethane solvent, and has the properties of mechanical stress discoloration and solvent mixing stress discoloration with high contrast and good repeatability. The solid powder of the rhodamine mechanochromic material containing a hydantoin structure has the characteristic of changing color upon grinding. After the action of mechanical force, the color of the powder changes from yellow to pink, and a reversible color change is achieved during the grinding-heating process.

The invention discloses a multi-section LCD liquid crystal screen icon dynamic driving device and a control method, and belongs to the technical field of liquid crystal display. Dynamic display is achieved, and the defect that a traditional broken code screen can only control display and closing of icons is overcome; a dynamic display technology which is only suitable for a high-end TFT screen originally is successfully transplanted to the low-cost VA field through an innovative driving method and hardware architecture design, and the blank of the technical field is filled; based on the VA screen, high fidelity which is not inferior to that of a high-end TFT is realized, the reduction degree of patterns and colors is high, and the cost is low. The invention has the advantages of low power consumption and high contrast: only microampere-level current is required at the minimum; the LED is still clearly visible under strong light, and is suitable for industrial instruments and vehicle-mounted display; the TN (twisted nematic) technology is adopted in the prior art, the visual angle and the contrast ratio are limited, the visual angle and the color performance are improved through the VA technology, the method is suitable for a color screen, and multi-color display is achieved through a color filter or backlight modulation.

The invention discloses an automatic calibration method and system for a cutting knife needle, a cutting machine and a storage medium. The method comprises the steps that a cutting knife needle is controlled to execute a gradient retraction cutting sequence, and the extension length of the knife needle retracts by a preset gradient length after each time of cutting; collecting a trace image formed on the calibration material by cutting each time through a camera; automatically identifying the shallowest effective cutting trace in the sequence based on image analysis; and calculating the reference zero point of the tool nose according to the sequence number corresponding to the trace, the preset gradient length and the initial extension length. The calibration material has a high-contrast substrate layer and contrast layer structure. The invention further discloses a system for implementing the method, the cutting machine comprising corresponding hardware and a storage medium. According to the invention, full automation of the calibration process is realized, dependence on artificial experience is eliminated, and the method has the advantages of high calibration precision, high efficiency, good consistency and the like.

According to the endoscope imaging fusion enhanced imaging method and device, imaging characteristics of white light (an overall structure background), red light (a deep blood vessel), green light (a high-contrast capillary vessel) and blue light (an enhanced surface microstructure) are comprehensively utilized, digestive tract mucosa can be analyzed in a layered mode, and comprehensive diagnosis and early focus recognition from the surface layer to the deep layer are achieved; during image fusion, adaptive weight configuration based on focus types is supported by adopting a weight dynamic adjustment mechanism, so that the quality of the finally fused image is influenced, and therefore, a neural network architecture is adopted to train each channel weight, and a more reasonable weight value can be obtained.

The invention discloses an industrial visual inspection method based on edge light scanning, which comprises the following steps of: initializing projector parameter setting, adjusting shooting frequency and exposure time of an industrial camera, acquiring defect images and forming an image sequence according to the size and the material of a detected piece and an expected defect type to be detected and based on an edge light scanning method; carrying out fusion operation on the image sequence, enhancing defect signals of the images through a weighted stacking algorithm by utilizing high-contrast characteristics of defect areas in multiple frames of images, and carrying out defect positioning and classification on the enhanced images by adopting an image segmentation technology, so as to realize visual detection on defects of the detected piece; by optimizing the design of an optical imaging system, edge light scanning of the quartz wafer is realized, defects are highlighted, surface defects can be quickly and accurately identified, the detection efficiency is greatly improved, defect characteristics are directly enhanced in an image acquisition stage, and quick and accurate defect detection is realized in combination with a sequence image analysis technology.