793 results about "Line scanning" patented technology

The invention relates to the technical field of industrial nondestructive testing and computer vision, and particularly discloses an intelligent detection method for surface defects of a container bottom plate based on deep learning. The method comprises the following steps: synchronously acquiring data through a laser radar and a line scanning camera, complementing a shielding area of a point cloud, and executing coordinate normalization to generate preprocessed data; constructing a lightweight feature alignment network to realize cross-modal feature mapping and pixel-level error correction; segmenting a defect area by adopting an improved PointNet + + network and reconstructing a three-dimensional model; extracting defect geometric parameters and combining with material attributes to perform stress simulation and life prediction; aggregating multi-port safety life data to construct a federated framework to update model parameters; and synthesizing physically real defect samples based on false detection cases, and injecting the defect samples into the network for training. According to the method, the bottleneck of missing detection of internal defects in traditional two-dimensional detection is overcome, the defect detection rate and quantification precision are remarkably improved, full-life-cycle safety evaluation of the container is supported, and the efficient requirement of automatic port inspection is met.

A 100-meter gridded spatialization method for carbon emissions of different land use types based on multi-source heterogeneous data is provided, including the following steps: S1, fitting Luojia-1A Satellite nighttime light data year by year with defense meteorological satellite program-operational linescan system (DMSP-OLS) and national polar-orbiting partnership-visible infrared imaging radiometer (NPP-VIIRS) fused 1-kilometer gridded nighttime light data, performing additive fusion on previous light data and current light data, introducing a time inertia weight factor to improve particle swarm optimization-back propagation (BP) neural network algorithm, and forming 100-meter nighttime light data correctable on a long time series; and S2, simulating a complex nonlinear dynamic changing relationship between multilevel 100-meter gridded data of different land use types in different industries and energy carbon emissions in different industries based on the 100-meter nighttime light data, and establishing a 100-meter spatialization inversion model for energy carbon emissions of different land use types in different industries.

The invention discloses a stainless steel tube surface defect detection method and system. The method comprises the following steps: S1, collecting stainless steel tube scanning images under dark field and bright field illumination; s2, splicing the line scanning images into dark field and bright field cylindrical expanded images, and executing brightness equalization and reflection suppression processing; s3, carrying out pixel-level fusion on the dark field and bright field cylindrical expansion images according to a set weight; s4, inputting the fusion cylinder expansion image into a YOLO network trunk of an integrated Swin Transform module, and extracting a multi-scale feature map; s5, performing feature fusion and bounding box prediction, and outputting a bounding box, confidence and a category label of the defect; s6, performing semantic segmentation, gray segmentation and weighted fusion in the detection frame area to generate a final defect mask; and S7, calculating the area, length, width and centroid coordinates of the defect, and converting the actual size and the spatial position. The stainless steel tube surface defect recognition accuracy and stability are improved.

The present disclosure relates to a line-scanning based super-resolution imaging method and apparatus. A line-scanning based super-resolution imaging method includes: providing a plurality of line-type structured excitation lights, each of the plurality of line-type structured excitation lights having a different illumination pattern; illuminating a target area in a line-by-line scanning manner along a single first direction while switching the plurality of line-type structured excitation lights, such that each row in the target area is illuminated by a corresponding one of the plurality of line-type structured excitation lights; and detecting response light generated by each row in the target area in response to being illuminated by the corresponding one of the plurality of line-type structured excitation lights.

The invention discloses a method for realizing high-quality stripping of large-size diamond by double-laser composite modification, which comprises the following steps: placing and fixing a sample on a three-dimensional mobile platform, and the upper surface of the sample is a crystal face (100); laser generated by an ultrafast laser generator penetrates through the upper surface of a sample and is focused into the sample, and then the laser generated by the ultrafast laser generator is used for performing first spot scanning on the sample to form an independent lattice structure; laser generated by the ultrafast laser generator is adopted to carry out secondary spot scanning on the sample to form a crack interconnected lattice structure; performing third-time line scanning on the sample by using laser generated by a nanosecond laser generator; and ultrasonically stripping the sample in water bath. By adopting a dual-laser composite modification process and combining bionic'honeycomb 'and'turtle shell' crack structure design, the crack is effectively guided to directionally extend along the crystal face (100), the problem of oblique crack when the stripping face is the crystal face (111) in the past is avoided, and the integrity and flatness of the stripping face are greatly improved.

The invention discloses a composite optical detection method and system for metal surface corrosion, and the method comprises the steps: employing a line scanning camera and a spectrum confocal sensor to carry out the synchronous data collection through a time-space registration method, obtaining the RGB image data of a to-be-detected metal surface through the line scanning camera, i.e., color data, and obtaining the color data of the to-be-detected metal surface; acquiring point cloud data, namely morphology data, of the metal surface to be detected by using a spectral confocal sensor; carrying out multi-modal data fusion on the RGB data and the point cloud data; a data fusion result is optimized by introducing a confidence coefficient weight; a double-threshold segmentation method of a color threshold and a depth threshold is adopted, and a corrosion area in an optimized data fusion result is preliminarily identified; performing morphological post-processing on the preliminary recognition result through opening operation and closing operation to generate a final recognition result; and carrying out quantitative evaluation on the corrosion area by utilizing a final identification result. The method provided by the invention can realize a rapid and accurate detection process.

The invention discloses an ultra-thin glass defect detection system and positioning method based on line scanning and phase deflection, and the system comprises a line scanning camera module, a distortionless phase deflection detection module, a high-magnification microscope module and a central processing unit. The line scanning camera module is used for collecting a 2D grayscale image of the ultrathin glass; the distortionless phase deflection technology detection module is used for collecting a surface height map of the ultra-thin glass; the high-magnification microscope module is used for carrying out depth positioning on the interlayer defect judged by the central processing unit; the central processing unit is used for receiving the 2D gray level image and the surface height map, executing an image registration and defect classification algorithm and sending a depth positioning control instruction to the high-magnification microscope module, a multi-module cooperative work detection system is constructed, the detection efficiency and the detection precision are both considered, and data support is provided for production process improvement.

The invention provides a steel plate surface defect detection method and system based on multi-source data fusion, and relates to the technical field of steel production automatic detection.The method comprises the steps that water stain pretreatment is conducted on the surface of a steel plate, the residual humidity is reduced to be below a preset low humidity threshold value, and a dry surface is obtained; based on the dry surface, collecting multi-source data through a 3D laser line scanning camera and a 2D industrial line scanning camera which are synchronously triggered, obtaining line scanning laser point cloud data, a line scanning reflectivity gray level image, a line scanning depth gray level image and an area array reflectivity color image, and establishing a space corresponding relation among the multi-source data; and forming a multi-source data set based on the spatial correspondence among the multi-source data. According to the method, a high-precision and high-robustness steel plate surface defect detection system covering the whole detection process is constructed by eliminating water stain interference, unifying a multi-source data reference, optimizing data quality, accurately identifying defects and realizing automatic feedback and data closed loop.

The invention discloses a visual labeling mechanism device and an automatic labeling method thereof, and the method comprises the steps: obtaining a workpiece surface image in motion in real time through a visual assembly, extracting a pose parameter of a workpiece in combination with an image processing algorithm, generating a pose data set containing a coordinate parameter and a deflection angle, and inputting a motion prediction model. Calculating a future movement track of the workpiece in combination with the speed parameter of the conveying line, and generating an adjustment instruction set of the labeling head; the movement track of the multi-axis mechanical arm is controlled based on the adjusting instruction set, the attaching pressure of the label contact face is adjusted through the pressure feedback module, and labeling operation of the curved surface of the workpiece is completed; a line scanning camera module is used for scanning the geometric position of the labeled label, a real-time detection result is generated by comparing the geometric position with a preset quality standard, a compensation signal is triggered if position deviation exists, and compensation operation is conducted through a compensation pressure head so as to ensure that the labeling quality meets the preset standard; and through cooperative control of visual identification, track prediction and compensation, the stability and the fitting precision of dynamic labeling are effectively improved.

The invention relates to a method for regulating and controlling crystallization of a perovskite thin film based on surface laser annealing and a preparation method of a perovskite photovoltaic module, and the surface laser annealing comprises the steps: constructing an infrared temperature measurement system in annealing equipment to monitor the surface temperature of the perovskite thin film in real time, and carrying out the closed-loop control of the output power of a surface laser; in annealing equipment, a preheating stage and a main annealing stage are executed through a surface laser, and optionally, a post-processing stage can be further added to prepare the perovskite thin film. Compared with the traditional heating stage annealing and line scanning laser annealing process, the method has the advantages that surface laser annealing is more uniform, local instantaneous heating is realized through high-energy dense surface laser within millisecond-second time, a solvent can be rapidly evaporated through rapid temperature rise, synchronous crystallization of a perovskite precursor is promoted, an independent precipitation time window of PbI2 is reduced, and the yield of PbI2 is improved. Therefore, the performance of the device is improved, and the high-efficiency perovskite solar cell module can be prepared.

The invention relates to the technical field of visual inspection, in particular to an intelligent pesticide finished product monitoring and conveying system based on industrial vision, which comprises a conveying mechanism for conveying pesticide finished products at equal intervals, and circulating clamping linear sweeping mechanisms symmetrically arranged on two sides of the conveying mechanism, a clamping linear scanning area can be formed between the two groups of circulating clamping linear scanning mechanisms; the circulating clamping linear sweeping mechanism comprises a circulating driving assembly and a clamping rotating linear sweeping assembly. The two sets of circulating clamping linear sweeping mechanisms are controlled to operate at the same speed as the conveying mechanism, when a pesticide finished product enters a clamping linear sweeping area, the pesticide finished product is clamped through the two sets of clamping rotating linear sweeping assemblies at the corresponding positions in the two sets of circulating clamping linear sweeping mechanisms, and the pesticide finished product moves in the clamping linear sweeping area along with the pesticide finished product; and meanwhile, the pesticide finished product is driven to rotate through the clamping and rotating line sweeping assembly during moving, and comprehensive line sweeping of peripheral information of the pesticide finished product can be achieved without stopping conveying of the pesticide finished product.

The invention discloses an industrial art material detection method based on multi-modal deep learning, particularly relates to the field of material analysis, and is used for solving the problems that cross-modal data alignment of curved surface utensils in highlight and multi-layer coating scenes is difficult, and material boundaries are easy to drift along with shooting batches and visual directions. The method comprises the following steps of: positioning a hyperspectral line scanning fragment; constructing a local reference grid to realize initial pairing; generating pixel-level residual image quantization dislocation; estimating luminosity mapping to form consistent bimodal data pairs after eliminating a high-reflectivity region; iteratively supplementing anchor points or adjusting grid rigidity, generating a stable material graph, updating an index table and a consistency record, ensuring cross-batch reusability and traceability, and improving the material detection precision.

The utility model discloses a waveguide slot antenna comprising a feed port layer, a slow wave line layer and a radiation slot layer which are stacked in sequence. The feed port layer is provided with a feed port used for inputting electromagnetic waves or outputting electromagnetic waves. The end face of one end, facing the feed port layer, of the slow wave line layer is provided with a snakelike slow wave cavity, one end of the snakelike slow wave cavity is provided with a feed signal end, and the feed signal end is correspondingly connected with the feed port. A plurality of waveguide radiation slots are formed in the snake-shaped channel of the snake-shaped slow wave cavity at intervals; a plurality of radiation cavity units in one-to-one correspondence with the waveguide radiation slots are arranged on the end face of one end, facing the slow wave line layer, of the radiation slot layer at intervals, a plurality of radiation slots are arranged in the radiation cavity units at intervals, and signal connecting ends correspondingly coupled with the waveguide radiation slots are arranged at one ends of the radiation cavity units. According to the utility model, the scanning angle of the antenna can be increased to adapt to application occasions with high antenna scanning angle requirements, and the adaptability is high.

The invention relates to the technical field of optical coherence elastography, in particular to a microscopic optical coherence elastography method and system.The method comprises the steps that a collection trigger signal is generated, a vibration control signal and a position control signal are output at the same time, and a target excitation system corresponding to a to-be-detected sample is determined; triggering the line scanning camera by collecting a trigger signal, collecting a line scanning image of the to-be-tested sample, and adjusting the position of the to-be-tested sample in the sample table so as to place the to-be-tested sample at a target test position; controlling a bending actuator or an ultrasonic transducer to vibrate by using the vibration control signal so as to excite low-frequency shear waves or ultrasonic waves in the sample to be detected; the to-be-detected sample is scanned through the position control signal to obtain propagation information of low-frequency shear waves or ultrasonic waves, and the pixel-level biomechanical properties of the to-be-detected sample are reconstructed by using a near-field full-wave inversion algorithm, so that the microstructure of a deeper tissue can be observed, the imaging speed is relatively high, and the detection performance of small inclusions is improved.

The invention discloses a fuel cell stack core straightness and perpendicularity detection device and method. A reference surface is positioned through a visual camera, and a transmission mechanism drives a laser line scanning instrument to perform line scanning point taking on an upper end plate and fit a reference plane; moving a visual camera through an X-axis or Y-axis transmission mechanism, scanning a long-edge extension line and a short-edge extension line of the upper end plate on a reference surface of the upper end plate along the horizontal direction, obtaining an intersection point, determining a coordinate origin, and performing reference determination by taking the long-edge extension line and the short-edge extension line as X-direction or Y-direction reference of a coordinate system; the transmission mechanism drives the laser line scanning instrument to carry out line scanning on the straightness of the side edge of the reactor core assembly, straightness line scanning data are obtained, line scanning point position coordinate values are obtained, data analysis is carried out according to the coordinate values, and compared with the prior art, high-precision detection of the straightness and perpendicularity of the fuel cell stack core is achieved, and the detection precision is improved. And the assembly quality and performance of the reactor core assembly are ensured.

The invention provides a pixel driving circuit and a display panel, and the pixel driving circuit comprises a driving unit which is used for driving a light-emitting element to emit light; the first switch unit is used for receiving a data signal; the second switch unit is used for controlling threshold voltage compensation; a first storage unit connected between the fourth node and the second node; the time sequence control circuit is used for respectively controlling the on and off of the first switch unit and the second switch unit so as to separate the data signal writing time from the compensation time, and controlling the Vth compensation time to be greater than the one-line scanning time; and in the light emitting stage, pulse type resetting is carried out on the source electrode potential of the driving unit by outputting a pulse signal. According to the pixel driving circuit provided by the invention, the technical problem that high and low refresh rates are difficult to be compatible in the prior art is effectively solved, the targets of high-quality display under the high refresh rate and low-power-consumption and flicker-free display under the low refresh rate are achieved, and the performance of display equipment and the user experience are remarkably improved.

The present invention relates to a coated cutting tool (1) for metal machining comprising a cemented carbide substrate body (5) and a coating (6) comprising a layer of a metal nitride, there is an uppermost part of the cemented carbide substrate body (5) wherein there are uppermost WC grains adjacent to the coating (6), there is an intermediate zone between an uppermost part of said WC grains and the coating (6) comprising a peak of Me3 intensity in an EDX line scan made from within the coating (6) into a WC grain perpendicular to the surface plane of the WC grain, Me3 is one of Ti, Cr, Zr, Mo, Nb and V, the average content of Me3 within + / - 10 nm from the Me3 max peak position is from 2 to 30 at%, the Me3 within + / - 1.0 nm from the Me3 max peak position in the EDX line scan is present as a metal nitride or carbonitride, the average content of any noble gas element or combination of noble gas elements, within + / - 1.0 nm from the Me3 max peak position in the EDX line scan, is ≤0.6 at%.

The present invention belongs to the technical field related to precision measurement, and discloses an orthogonal line spectrum confocal scanning profile measurement sensor and an anti-vibration method. The sensor includes a light source module, and first and second direction linear dispersion confocal imaging modules. The light source module is used to generate two beams of orthogonal collimated white light. The two beams of orthogonal collimated white light enter the first and second direction linear dispersion confocal imaging modules respectively, thereby scanning the surface of the sample to be measured and imaging, thereby obtaining the surface profile of the sample to be measured at two orthogonal angles. The first direction linear dispersion confocal imaging module includes a first linear dispersion confocal structure, a second slit, and a first imaging unit. The second direction linear dispersion confocal imaging module includes a second linear dispersion confocal structure, a third slit, and a second imaging unit. The directions of the second and third slits are respectively the same as the directions of the two orthogonal slits in the first slit. Through the present invention, the problem of vibration affecting measurement accuracy during line scanning surface measurement is solved.

The invention discloses an electron beam composite scanning filling melting forming method and device, and relates to the technical field of additive manufacturing. A melting forming method for electron beam composite scanning filling comprises the following steps that a three-dimensional digital model of a part to be printed is established and sliced and layered, and layered scanning data are obtained; dividing a contour edge area and an internal area of the section of the part based on the hierarchical scanning data; based on a random discrete point algorithm, generating a random discrete point scanning path of the contour edge area; generating a continuous line scanning filling path of the internal region based on a parallel equidistant line filling method; in the contour edge area, according to a random discrete spot scanning path, a pulsed electron beam is adopted for spot scanning, filling and melting; and in the internal area, according to a continuous line scanning filling path, a focused electron beam is adopted for line scanning filling melting, single-layer filling melting is completed, and then filling melting is completed layer by layer till the part is formed. According to the method, collaborative optimization of the surface quality and the printing efficiency is achieved, and the problem that at the present stage, the surface roughness of an electron beam powder bed printing part is too high is solved.

The invention discloses a rotating mirror scanning light path system for generating point ring light and a control method thereof. The light path system comprises a light source module used for outputting laser; the shaping and focusing module comprises a rotary reflecting mirror and an arc-shaped reflecting mirror which are sequentially positioned on a laser light path; the rotating reflector comprises a central area and an outer ring area, the outer ring area surrounds the central area, and the curvature radiuses of the central area and the outer ring area are different; the rotary reflecting mirror is used for reflecting the laser and shaping the laser into a point annular light spot; the arc-shaped reflector is used for focusing the annular spot; the control module is used for controlling the rotating reflector to rotate around the rotating shaft, and laser scanning in the first direction is achieved. And the control module is also used for controlling the rotary reflector and the arc reflector to integrally move and rotate to realize line scanning and rotary scanning of the laser in the second direction. According to the laser shaping device, only two arc-shaped optical reflectors and one optical fiber output laser are needed, laser can be shaped into point annular light spots, splashing can be reduced, the porosity can be reduced, and the machining quality can be improved.

The utility model discloses a three-dimensional scanner. The three-dimensional scanner comprises a photoelectric support, a laser emitting assembly and an acquisition module. The front end of the photoelectric support is provided with a first mounting hole, a second mounting hole and a third mounting hole, and the rear end is provided with a first mounting cavity. The laser emission assembly comprises a speckle laser emission module, a line laser emission module and an installation support, the installation support is installed in the first installation cavity, and the speckle laser emission module and the line laser emission module are installed on the installation support and are exposed through the first installation hole and the second installation hole respectively; the speckle laser emission module and the line laser emission module are respectively used for emitting speckle laser and line laser to a target; the acquisition module is mounted in the third mounting hole and is used for acquiring speckle laser and line laser reflected by the target; the mainboard is installed at the rear end of the photoelectric support and is electrically connected with the speckle laser emission module, the line laser emission module and the acquisition module. The three-dimensional scanner integrates two scanning modes of line scanning and speckle scanning.

The invention provides a full-section detection method, device and equipment for reinforcing a steel rail welding seam and a storage medium, and relates to the technical field of nondestructive testing. The method comprises the following steps: respectively arranging a plurality of first phased array probes on rail head treads and rail bottom surfaces of a first steel rail and a second steel rail, and arranging a second phased array probe right above a welding seam; the detection method comprises the following steps: determining optimal parameters of a transducer based on the material of the steel rail, the size of the cross section of the steel rail and the sound pressure intensity at a preset focusing depth position; determining the optimal scanning angle range of the first phased array probe based on the propagation velocity of the ultrasonic longitudinal wave in the wedge block and the steel rail, the propagation velocity of the ultrasonic transverse wave in the steel rail and the Snell's law; and controlling the first phased array probe to carry out sector scanning imaging within the optimal scanning angle range, and controlling the second phased array probe to carry out zero-degree line scanning imaging, so as to detect the full section of the welding seam. According to the invention, the welding seam can be accurately detected without dismounting the clamping plate.

The invention relates to the technical field of PTFE (polytetrafluoroethylene) membrane micropore defect detection, in particular to PTFE membrane micropore defect detection equipment based on machine vision, which comprises symmetrically arranged mounting side plates and sensor assemblies arranged on one sides of the mounting side plates at intervals, and an overturning assembly used for realizing driving is arranged between the two groups of mounting side plates. Adjusting assemblies are arranged on the outer side of the overturning assembly in an array mode, and detection modules used for detecting PTFE membrane micropore defects are arranged on the outer sides of the adjusting assemblies. And the adjusting assembly comprises clamping frames arrayed on the outer side of the overturning assembly, bent plates are symmetrically arranged on the sides, close to the overturning assembly, of the clamping frames, and quick connecting rods are installed on the inner walls of the bent plates at intervals up and down. The device has the beneficial effects that high-resolution imaging is carried out by adopting the intelligent sensor, the white light interferometer and the high-speed line scanning camera, micro-current detection is realized in combination with weak current released by the discharge piece, micropore defects can be accurately identified, and misjudgment or missing detection caused by a single detection mode is avoided.

The embodiment of the invention provides a vector file data processing method, a vector file data processing system and related equipment, which are used for improving the lattice processing efficiency of vector files. The method in the embodiment comprises the following steps: reading a Gerber file, and sorting all primitives according to a drawing sequence; boundary fusion processing is conducted on the adjacent primitives, and the boundary fusion processing comprises the steps that whether boundaries between the adjacent primitives coincide or not is judged, and if yes, the boundaries are merged to generate new primitives; and partitioning the primitives subjected to boundary fusion processing to form a plurality of data blocks, and performing multi-thread line scanning processing on the primitives in the data blocks to obtain a dot matrix file.

The invention discloses a chip packaging body defect detection method, and relates to the field of chip detection.The chip packaging body defect detection method comprises the steps that a feeding position, a visible light detection position, a material connection position, an infrared light detection position and a discharging position are arranged on a rack side by side; the first conveying module lifts the carrier plate at the visible light detection position so that the scanning head can carry out line scanning imaging on the carrier plate units, the second conveying module lifts the carrier plate at the infrared light detection position so that the infrared system can carry out imaging, and physical separation and process relay of visible light and infrared detection are achieved. On the basis, position offset and surface defects are obtained based on visible light images of the carrier plate units, internal subfissures are recognized in combination with infrared images of the chip bodies, and unit-level judgment is generated by integrating multi-source results, so that the efficiency bottleneck of a traditional start-stop mode is avoided, and the coverage blind area of single-mode detection is overcome; and the takt, the integrity and the accuracy of chip packaging body defect detection under a high-productivity production line are remarkably improved.

A visual image processing method based on brain-like processing belongs to the technical field of image recognition, and comprises the following steps: step 1, dividing an image into multiple rows and columns by taking pixels as units, scanning row by row, and obtaining digital visual waves of the rows by taking a pixel value of each point as a vertical coordinate and the number of the columns as a horizontal coordinate; step 2, screening out a to-be-screened wave band of each row; step 3, taking the central point of each to-be-screened wave band, and screening the front end point and the rear end point; and step 32, carrying out graph fitting identification on the central point and the end point of the to-be-screened wave band reserved in each row and each column to obtain an image. According to the method, the image pixel value change points are directly subjected to extraction reasoning and fitting clustering in a line-by-line scanning and operation mode to obtain and recognize the image, the image recognition speed and precision are improved, the hardware requirement is reduced, and the method is suitable for performing real-time automatic search recognition and tracking application on moving and small dim targets in various complex environments.

The invention discloses a coherent interference-based coaxial transceiving line scanning three-dimensional profile measurement system and method. The system comprises a low-coherence light source; the beam splitter is used for collimating light emitted by the low-coherence light source and then dividing the light into two paths; the measuring light is focused into a linear light beam through the cylindrical lens and is emitted to the surface of the to-be-measured sample, and the measuring light reflected by the surface of the to-be-measured sample returns to the beam splitter along the original path; the reference light is emitted to a reference mirror with a certain inclination angle and is reflected back to the beam splitter through the reference mirror; the measuring light reflected by the surface of the sample to be measured and the reference light reflected by the reference mirror are subjected to beam combination interference through a beam splitter and are imaged to enter a camera for interference pattern acquisition; the signal processing system is used for processing the interference patterns to obtain height distribution of the to-be-measured sample on a line field, and three-dimensional profile measurement of the to-be-measured sample is realized by transversely moving or transversely scanning the to-be-measured sample and overlapping the interference patterns measured at different positions. Therefore, high-efficiency measurement can be realized while the micro-nano precision is kept.

The method comprises the following steps that two-dimensional image data and three-dimensional point cloud data of a to-be-detected product are synchronously collected through a multi-mode sensor, and the multi-mode sensor comprises a high-resolution line scanning camera, a TOF sensor and a tunable annular light source; and matching a multi-brand knowledge base based on the product identifier or the appearance feature, and calling a feature template and a defect judgment standard of a corresponding brand. According to the visual detection method and system capable of self-adapting to the multiple brands, through the transfer learning framework and the multi-brand knowledge base, new brand products can be automatically identified, the pre-training model can be loaded, high initial detection accuracy can be achieved without extra annotation data, and by combining the self-supervised generative adversarial network and the federal learning mechanism, the detection accuracy is improved. And when the defect sample size is small, the relatively high defect detection rate can still be realized, and the false alarm rate is relatively low.

The invention provides a calibration method suitable for a single-camera high-precision scanning system. The calibration method comprises the following steps: S1, carrying out area array mode calibration on a linear array camera; by adjusting the exposure time of the line-scan digital camera, moving the base station to image at different positions of the checkerboard, and solving internal parameters and distortion parameters of the line-scan digital camera; s2, line scanning mode calibration; performing line scanning imaging on the checkerboard to obtain pixel coordinates of angular points of the checkerboard; and separating the actual rotation angle of the checkerboard from the projection error, and further deducing the rotation angle of the real camera, thereby completing calibration. According to the calibration requirement of the linear array camera in a scanning imaging scene, a two-stage algorithm combining area array mode modeling and linear scanning imaging correction is provided, and the problem that a traditional calibration method is not applicable under the condition that a single camera is fixed and moves in a single direction is effectively solved.

The invention discloses an in-situ nondestructive measurement method, medium and equipment for elastic parameters of a multi-thin-layer heterogeneous material, and belongs to the technical field of ultrasonic measurement. The method comprises the steps that a multi-mode ultrasonic propagation multi-parameter model is established for a to-be-measured structure, a propagation time theoretical value corresponding to each ultrasonic mode is calculated, and the to-be-measured structure is a multi-thin-layer heterostructure; laser ultrasonic linear scanning is carried out on the to-be-detected structure, ultrasonic signals of all ultrasonic modes propagating in different directions are collected, and a line scanning ultrasonic data set is obtained; for a selected ultrasonic mode, extracting propagation time experimental values under different excitation-receiving point combinations from ultrasonic signals of the line scanning ultrasonic data set; and aiming at the selected ultrasonic mode, minimizing the difference between the propagation time theoretical value and the propagation time experimental value as a target function, and using a parameter optimization algorithm to invert the elastic parameter of the to-be-tested structure. According to the invention, long-distance and non-contact in-situ nondestructive measurement can be realized, and the operation is simple.