939 results about "High dynamic range" patented technology

The invention relates to the technical field of optical detection, in particular to an automatic optical detection method and system based on gasket surface defects. The method comprises the following steps: carrying out parameter extraction on a to-be-detected gasket, obtaining inner and outer diameters, thickness and tolerance range information of the gasket, and obtaining gasket structure parameter data; matching, calibrating and storing the gasket structure parameter data according to a preset element parameter library to obtain a gasket element parameter library; performing surface scanning on the to-be-detected gasket by using a visual sensor to generate original image data; performing multi-channel optical imaging processing on the original image data, and performing illumination unevenness compensation by using a high dynamic range technology to obtain enhanced surface image data; and carrying out image preprocessing on the enhanced surface image data to obtain gasket surface detection image data. According to the method, the multi-channel optical imaging and advanced image processing technologies are adopted, the self-adaptive algorithm is combined, tiny defects can be accurately recognized, and the false detection rate and the omission ratio are remarkably reduced.

The invention belongs to the technical field of industrial measurement, and discloses a pump machine metal shell size detection method based on image analysis, which comprises the following steps: synchronously acquiring image data of a pump machine metal shell in a visible light wave band and a near-infrared wave band; the contrast ratio of image data is optimized through a dynamic exposure control technology, and a high-dynamic-range multispectral image is obtained; three-dimensional geometric information of the pump machine metal shell is obtained, registration fusion is carried out on the three-dimensional geometric information and the high-dynamic-range multispectral image, a dense three-dimensional point cloud model with multispectral textures is constructed, and a multi-frequency heterodyne algorithm is adopted to process a high-reflection area in the dense three-dimensional point cloud model; the method comprises the following steps: collecting a multi-angle reflection image of a pump machine metal shell by rotating a linear polarizer, calculating a Stokes vector to extract a diffuse reflection component, carrying out diffuse reflection component constraint and brightness suppression on a high-reflection area based on a CIE-Lab color space, and outputting a dense three-dimensional point cloud model after texture enhancement; and the size detection precision of the pump machine metal shell is improved.

The invention provides a virtual reality building roaming method and system based on scene rendering, and relates to the technical field of virtual reality, and the method comprises the steps: dividing building three-dimensional model data through an octree space segmentation algorithm, and determining a detail level threshold according to a viewpoint distance, and constructing a multi-level detail model. And baking the basic illumination chartlet by using a pre-calculation irradiance transmission technology, and performing real-time synthesis in combination with the multi-resolution shadow chartlet. And calculating illumination by adopting a physics-based rendering method, and fusing the illumination with the scene light and shadow synthesis data to generate dynamic illumination rendering data. And determining a current viewpoint based on the spatial position and attitude data of the virtual reality head-mounted display, performing view cone elimination and shielding elimination, and performing GPU instantiation rendering on visible building data. And the rendered image is optimized through high dynamic range processing and a time domain anti-aliasing technology, and a final scene image is generated in combination with an atmospheric scattering algorithm. According to the invention, the rendering efficiency and scene reality sense of virtual reality building roaming can be effectively improved.

The invention relates to a method and a system for inhibiting strong mutual reflection based on polarization complementary logic coding. The method comprises the following steps: acquiring calibration parameters of a camera and a projector; generating a polarization phase shift stripe based on the calibration parameter, designing a high-frequency complementary logic code word, and embedding the high-frequency complementary logic code word into a polarization channel through polarization modulation to obtain a composite stripe pattern; the method comprises the following steps: calculating to obtain a linear polarization degree image through an acquired multi-angle polarization image, extracting a high-reflection region from the linear polarization degree image by adopting an adaptive threshold segmentation method, and carrying out binarization processing on the linear polarization degree image to obtain a binarized linear polarization degree mask; and on the basis of a decoding range of a binary linear polarization degree mask constraint high-reflection region, decoding the composite stripe pattern to obtain absolute phase level information, and performing noise suppression on the absolute phase level information by adopting a self-adaptive threshold algorithm. The method can effectively solve the influence of mutual reflection and high dynamic range of the surfaces of the measured object, and improves the precision and stability of three-dimensional reconstruction.

The invention relates to the technical field of laser control, and provides a high-dynamic-range multi-channel laser interference control method and related equipment. Channel signals of a plurality of target channels are obtained in real time according to a plurality of preset photoelectric detectors, time domain-frequency domain-space three-dimensional sampling is performed on the channel signals to construct an original tensor, joint optimization processing is performed on the original tensor to obtain a noise reduction tensor, phase unwrapping is performed on the noise reduction tensor to obtain an absolute phase matrix, and the absolute phase matrix is subjected to noise reduction processing to obtain an absolute phase matrix. Performing dynamic range expansion on the absolute phase matrix to obtain an enhanced phase matrix, performing multi-wavelength calculation and environment compensation on the enhanced phase matrix to obtain a distance matrix, and performing vibration compensation on the distance matrix to obtain a piezoelectric driving signal. According to the channel laser interference control method, through a series of advanced algorithms and processing means such as multi-dimensional signal acquisition, combined noise reduction optimization, precise phase unwrapping, dynamic range expansion and vibration compensation, comprehensive improvement of laser interference measurement signals is realized.

The invention discloses a three-dimensional reconstruction method and system based on a line shift coding pattern, and the method comprises the steps: designing an original fringe pattern in a line shift manner to obtain a plurality of coding patterns, and carrying out the projection and collection of a to-be-measured object through the plurality of coding patterns and a preset Gray code pattern, respectively obtaining a plurality of fringe projection drawings and a plurality of Gray code projection drawings; synthesizing M phase shift fringe patterns according to the collected multiple fringe projection patterns; performing phase calculation on the M phase shift fringe diagrams to obtain a folded phase diagram; and combining the folded phase diagram with the plurality of Gray code projection drawings for phase unwrapping to obtain an unwrapped phase diagram, and reconstructing the to-be-measured object by using a three-dimensional reconstruction principle. According to the method, fringe synthesis is firstly carried out on the plurality of coding patterns, and then phase calculation is carried out on the M phase shift fringe patterns, so that three-dimensional reconstruction is carried out on the to-be-measured object by using the phase, relatively high reconstruction precision is achieved, the robustness is relatively high, and the method has a relatively good application effect in a high dynamic range scene.

In some embodiments, an electronic device generates an automatically-generated visual media using one or more recognized concepts extracted from a prompt inputted by a user. The recognized concepts include personalized template subjects and / or prompt suggestions. While displaying the user interface including the recognized concepts, the electronic device receives one or more inputs to modify the recognized concepts. The electronic device generates multiple variants of the automatically-generated visual content using the one or more recognized concepts. The electronic device adds an automatically-generated visual content to a content entry field of an application, different than the automatically-generated visual media application, without opening the automatically-generated visual media application. The electronic device applies a visual effect to content that is generated using an artificial intelligence model. The electronic device displays visual information corresponding to an artificial intelligence model. The electronic device displays an animation including displaying a user interface with high dynamic range luminance.

A system, method, and computer program product are provided for high-dynamic range images. In use, a first pixel attribute of a first pixel is received and a second pixel attribute of a second pixel is received. Next, a scalar based on the first pixel attribute and the second pixel attribute is identified. Finally, the first pixel and the second pixel are blended, based on the scalar, wherein the first pixel is brighter than the second pixel. Additional systems, methods, and computer program products are also presented.

The invention discloses a metal processing detection system and method based on computer vision, and relates to the technical field of defect detection.Rotatable polarization filters are installed in front of a lens of a vision sensor, the polarization filters are arranged at four angles respectively, and four sets of polarization images are collected; calculating the polarization degree of each pixel point on the surface during metal processing by using the gray values of the four groups of polarization images, setting a polarization region interval, and dividing a metal surface reflection region into a specular reflection region and a diffuse reflection region; acquiring images of different wavebands by using a multispectral vision sensor, and constructing a multispectral image set; for the same wave band, the images of different exposure times are synthesized to obtain an HDR high dynamic range image; setting a weight for the HDR image of each wave band, and combining the weighting of the HDR image and the suppression of the polarization degree to carry out multispectral image fusion so as to obtain a final image; and a gray threshold value and an area threshold value are respectively calculated through the average value and the standard deviation, and the metal processing defects are judged by using the two threshold values.

The invention provides a glass wine bottle surface defect detection system. The glass wine bottle surface defect detection system comprises the following steps: S1, carrying out image acquisition; s2, a step of image processing; s3, a step of defect identification and classification; the step S1 specifically comprises the following steps: S11: selecting and configuring equipment, selecting an industrial-grade high-definition camera which has the characteristics of high resolution, low noise and high dynamic range according to the size, the shape and the surface characteristics of the glass wine bottle so as to ensure that a clear image is captured, and configuring an LED backlight or annular light source system so as to eliminate shadow and light reflection, so that the glass wine bottle is protected; details on the surface of the wine bottle are highlighted, and brightness and color temperature of the light source are adjusted according to scenes; according to the method, the problems of low efficiency and low detection accuracy caused by experience judgment or manual visual observation during existing glass wine bottle detection are well solved, and the intelligent detection method based on visual observation is provided to improve the detection efficiency and accuracy.

The invention relates to the technical field of three-dimensional model design of molds, and provides a mold three-dimensional high-precision modeling method and system based on multi-modal data fusion, and the method comprises the steps: collecting an RGB image, a high dynamic range image and a structured light depth image of an entity mold, and carrying out the reconstruction to obtain a globally consistent dense point cloud; denoising and semantic segmentation are carried out on the point cloud, and a cavity function area, a core function area and a runner function area are recognized; performing differential parametric modeling based on an identification result: performing geometric primitive fitting on regular geometric features, performing NURBS curved surface reconstruction on a free-form surface, and performing swept volume or rotator fitting on a flow channel to obtain a parameterized geometry, a parameterized curved surface and a parameterized channel model; and applying boundary continuity constraint to fuse the parameterized geometry, the parameterized curved surface and the parameterized channel model, and constructing a hybrid parameterized three-dimensional model. According to the method, rapid reconstruction of a high-precision, high-semantic and full-parameterized three-dimensional model of an entity mold can be realized.

The invention discloses a die casting burr detection system and method based on machine vision, and relates to the field of machine vision detection.The method comprises the steps that based on a spectral feature vector set, a curved surface self-adaptive optical compensation mechanism is used, a polarization-spectral image set is obtained again through light path optimization, and the optimized polarization-spectral image set is obtained and corrected; generating a high-resolution image set, performing illumination distribution analysis on the surface of the die casting through the high-resolution image set to form a multi-exposure image set, inputting the multi-exposure image set into the constructed U-Net model for semantic segmentation, generating a high-dynamic-range image, and performing shadow elimination on the high-dynamic-range image by applying a gradient Poisson fusion algorithm to obtain a high-dynamic-range image. And after the shadow is eliminated, a pre-trained Mask-R-CNN model is used for detection, and a burr detection result is output. According to the method, the polarization-spectrum image set is optimized through curved surface adaptive optics, and imaging optimization and geometric correction of a complex curved surface area are achieved.

For a high dynamic range surface, the invention provides a simple and effective three-dimensional reconstruction method, which combines the characteristics of non-overexposure areas in a polarization image and a non-polarization image, and aims to improve the reconstruction quality of a high dynamic range surface object. A non-overexposure area in a non-polarization picture generally contains complete stripe information and is less influenced by highlight interference. The optimal stripe regions of different polarization picture groups are extracted, then the incomplete and unclear stripe information in the non-polarization image is gradually replaced by the optimal stripe information in different polarization directions, and finally, the stripes in the overexposure region in the non-polarization image are replaced by the complete stripe information. In this way, global complete stripe information can be obtained. Experimental results show that when three-dimensional reconstruction is carried out on an object with a strong reflection surface, the negative influence of an overexposure area on absolute phase calculation and three-dimensional reconstruction can be effectively solved.

The invention relates to the technical field of tunnel lighting, in particular to a tunnel portal lighting control method and system, and the method comprises the steps: S1, obtaining a first image in a short exposure time mode; obtaining a second image in a long exposure mode; s2, synthesizing the second image and the first image into a high-dynamic-range image, and dividing the high-dynamic-range image into a low-brightness area, a transition area and a high-brightness area; s3, generating a brightness transition curve of the transition area; s4, determining a brightness adjusting curve of the transition area according to the maximum brightness of the tunnel portal and the light and shade adaptive pupil area change rate; s5, determining the target brightness of each part of the transition area according to the brightness transition curve and the brightness adjustment curve; and S6, adjusting the brightness of the illumination lamp in the transition area to enable the brightness of each part of the transition area to reach the target brightness. The target brightness is determined according to the brightness transition curve in combination with the brightness adjustment curve, the influence of the brightness adaptive pupil area change rate on the sight line is fully considered, and the change of the brightness of the tunnel portal is more adaptive to the function of human eyes.

The invention relates to an event-based lip reading method and device, equipment and a storage medium. The method comprises the following steps: collecting an original event stream of a lip sequence image through an event camera; therefore, the brightness change of each pixel can be asynchronously recorded with microsecond-level time resolution, and ultra-low delay, high dynamic range and sparse data representation are realized. Converting the original event stream into a frame-shaped event tensor based on a voxel representation method to obtain a voxelized event body; performing spatial feature extraction on the voxelized event body through a front-end network in an event-based lip reading model to obtain multi-scale spatial features; performing time-dependent modeling on the multi-scale spatial features through a rear-end sequence model in an event-based lip reading model to obtain a sequence code; therefore, space and time features can be fused, and the accuracy and stability of sequence coding are improved. And determining the lip reading recognition content according to the sequence code. Therefore, stable recognition of lip movement can be realized, and lip reading accuracy is improved.

The invention provides an aerial image detail enhancement method based on deep learning. The aerial image detail enhancement method comprises the following steps: constructing a cyclic convolutional neural network with a memory state; the cyclic convolutional neural network with the memory state comprises a denoising network, an event reconstruction image network and a fusion network; in a low-illumination scene, a low-quality video frame sequence VL * and a low-quality event stream FL * of a target area are acquired respectively in each acquisition period through a visual camera and a common camera, the low-quality video frame sequence VL * and the low-quality event stream FL * are input into the trained cyclic convolutional neural network with the memory state, a reconstructed first high-quality video frame sequence VHR1 * is output, and a reconstructed second high-quality video frame sequence VHR2 * is output; the video frame sequence is a finally obtained video frame sequence after image detail enhancement. According to the model designed by the invention, the high dynamic range information of the event stream data and the high texture detail information of the image frame sequence data are utilized at the same time, and the visual perception enhanced image with the high dynamic range and the high texture detail information at the same time can be constructed through fusion.

Disclosed in the present invention is a high-dynamic capacitive digital microphone system based on electrostatic force feedback. The system comprises an MEMS microphone and an interface circuit chip thereof, wherein the interface circuit chip comprises a charge pump, an analog front end, an analog-to-digital conversion circuit, an amplitude detection module, an amplitude adjustment control module and a pulse density modulation module. In the present invention, the characteristic of the sensitivity being able to be adjusted by means of changing a bias voltage by a capacitive sensor is used, and a charge pump with an adjustable output voltage is used to provide electrostatic force feedback, thereby further expanding the adjustable range of a front-end gain, and also reducing requirements for the noise performance of a readout circuit. Therefore, the system in the present application has the characteristics of high dynamic range and high-precision quantization.

The invention discloses an electric power cross-operation dynamic risk prediction and early warning method, system, equipment and medium based on multi-modal data fusion, and relates to the technical field of intelligent electric power systems, and the method comprises the steps: collecting multi-source data of a cross-operation site in real time, including personnel spatial position data, equipment state data, environmental parameters and operation flow information; performing multi-modal fusion processing on the multi-source data, and constructing space-time risk association features; constructing a dynamic risk factor matrix, performing fusion calculation on the static risk reference value and the dynamic correction value, and dynamically updating the weight of each factor along with time; predicting a risk change trend based on the combination of a time sequence prediction model and a graph structure modeling method; and generating a three-dimensional visual risk thermodynamic diagram based on a prediction result, and triggering a multi-level early warning mechanism when a risk value exceeds a threshold value. The image is reconstructed through multi-modal fusion processing, and key edge features can be recovered in complex environments such as low illumination, high dynamic range and fast motion.

The invention discloses an HDR video reconstruction method based on a standardized stream, and belongs to the technical field of high dynamic range image processing. The method comprises the following steps of: firstly, constructing a convolution optical flow estimation module with a self-adaptive normalized structure, wherein the convolution optical flow estimation module is used for accurately acquiring optical flow information between adjacent frames in an alternative exposure LDR video image sequence; then, carrying out multi-level feature alignment on the image sequence through an image alignment module so as to reduce alignment errors caused by illumination difference and movement; and finally, inputting the aligned and fused multi-level LDR image features into a standardized flow reconstruction network to realize high-quality HDR video image reconstruction. Aiming at the video reconstruction problem under the alternate exposure condition, the invention designs a standardized flow modeling structure considering the optical flow estimation precision and the feature alignment effect, and effectively improves the HDR video reconstruction quality in a complex dynamic scene.

The invention relates to a measuring range dynamic adjusting method based on amplitude parallel estimation, which comprises the following steps of: (1) acquiring N current signal sampling points at a sampling rate fs, and calculating the amplitudes of the maximum N / 4 sampling points in the N sampling points in parallel; (2) storing N / 4 sampling points and the corresponding measuring range interval A; (3) continuously carrying out k times of signal sampling to obtain k groups of range interval sets {A1, A2,..., Ak}, and selecting the latest four groups of range intervals to carry out combined judgment; (4) if the four groups of measuring range intervals are consistent, keeping the current measuring range; if a wide-range interval occurs in the last two groups, switching to a wide range; and (5) carrying out trend estimation based on nearly five actual measuring range intervals. According to the invention, the key problem of the intelligent monitoring system in a high dynamic range signal monitoring scene is solved, the intelligent monitoring system can accurately judge the signal range and quickly adjust the range, and the loss of effective signals is reduced.

The invention discloses a method and a system for detecting residues in a closed steel drum based on visual inspection, and belongs to the technical field of industrial nondestructive inspection and computer vision crossing, and the method comprises the following steps: collecting a multi-angle image through dynamic light field regulation and control, and generating a high dynamic range image; constructing a dense three-dimensional point cloud by applying a semi-global matching algorithm, and fusing the infrared heat absorption coefficient and the visible light texture roughness to generate a three-dimensional model with material attributes; based on structural features, filtering out potential targets in a non-residual region, and combining point cloud height variance and HSV hue gradient analysis to generate matter state discrimination parameters; training a convolutional neural network by using the voxelization model and the circumferential weld feature template, and outputting an anti-interference residue segmentation mask; a two-dimensional projection drawing is obtained through cylindrical coordinate expansion, and the residue weight is calculated by combining a density mapping library. According to the method, a dynamic light field regulation and control and multispectral fusion technology is adopted, and a three-dimensional point cloud reconstruction and deep learning algorithm is combined, so that accurate detection and quantitative analysis of the residues in the closed steel drum can be realized.

To obtain a practical yet versatile and accurate apparatus for luminance re-grading (300) of an input high dynamic range image (IM_HDR) of first luminance dynamic range into a second image (IM_DR2) of second luminance dynamic range, the inventors propose the apparatus for luminance re-grading (300) of an input high dynamic range image (IM_HDR) of first luminance dynamic range into a second image (IM_DR2) of second luminance dynamic range, wherein a maximum luminance of the second image may be higher or lower than a maximum luminance of the input high dynamic range image, the apparatus comprising:—a first neural network processing circuit (301), having as input a set of pixel luminances of the input high dynamic range image (IM_HDR), wherein the first neural network processing circuit has at least two sets of outputs (S1, S2), wherein the sets of outputs have output nodes for supplying parameter values of a parametric re-grading equation, wherein the parametric re-grading equation is composed of at least two partial functions an amount of which in the re-grading equation being controlled by at least two corresponding output parameter values (P11, P21); wherein a first output set (S1) comprises at least two parameter values (P11, P21) which determine a first shape of the re-grading function, and a second output set (S2) comprises at least two corresponding parameter values (P12, P22) which determine an alternative second shape of the re-grading function, which being determined by the same parametric definition but using different values for its parameters gets a different shape;—a second neural network processing circuit (302), which has as input at least one measurement value (La) from at least one sensor (311), and as output nodes supplying a set of weights corresponding in amount to at least the number of sets of outputs of the first neural network processing circuit;—a combiner (303), arranged to add a first parameter value (P11) of the first output set of the first neural network processing circuit multiplied by a corresponding first weight (w11) from the second neural network processing circuit, to a corresponding parameter value (P12) for the same re-grading function defining parameter of the second output set of the first neural network processing circuit multiplied by a corresponding second weight (w12) from the second neural network processing circuit, yielding a final parameter value (P1_f); and—a luma mapping circuit (320) arranged to map input lumas of the input high dynamic range image (IM_HDR) with a luma mapping function (F_L) which is defined by at least two parameters comprising the final parameter value (P1_f) to obtain output lumas of the second image (IM_DR2).

An image sensor may include an array of imaging pixels arranged in rows and columns. Each imaging pixel may include a photodiode, an overflow capacitor, an overflow transistor that is interposed between the photodiode and the overflow capacitor, a floating diffusion region, a transfer transistor that is interposed between the photodiode and the floating diffusion region, a voltage supply, and a reset transistor that is interposed between the floating diffusion region and the voltage supply. The voltage supply may provide a voltage at a first magnitude that is less than the pinning voltage for a first portion of a reset period and may provide the voltage at a second magnitude that is greater than the pinning voltage for a second portion of the reset period.

The utility model describes a pixel circuit and a pixel array of an image sensor, and the pixel circuit comprises a first photosensitive unit which is configured to receive an optical signal in a first exposure duration so as to generate a first image signal; a first floating diffusion node connected to the first photosensitive unit; the first transfer transistor is connected between the first photosensitive unit and the first floating diffusion node; the gain control circuit is connected with the first floating diffusion node so as to collect a second image signal generated under a second exposure time length and collect a third image signal generated under a third exposure time length, and the second exposure time length is smaller than the first exposure time length and smaller than the third exposure time length; the gain control circuit at least stores the third image signal in the overflow unit and then outputs the third image signal; the reset circuit is connected between the gain control circuit and the first power supply voltage; the readout circuit is connected to the first floating diffusion node. According to the pixel circuit, the dynamic range can be enlarged, and the problem of motion artifacts is solved.

The invention relates to the technical field of LED display screens, and discloses a visual interaction method, device and equipment for an LED display screen and a storage medium, and the method comprises the steps: carrying out the illumination balance and group posture recognition of image data of an audience region, and obtaining the thermodynamic diagram data of the group posture of the audience; performing high dynamic range feature extraction and two-way visual language attention decoding on an LED display screen picture to obtain display content semantic region data; performing manifold distance audience behavior clustering and view angle difference influence processing to obtain display response trigger data; performing potential field model and model prediction control on the display parameters to obtain a display parameter control instruction; performing interaction effect analysis to obtain a display screen parameter optimization strategy; according to the method, the interaction intelligence level of the LED display screen is remarkably improved, and the technical problem of group behavior recognition of a traditional method is effectively solved.

The present disclosure relates to a pixel circuit of an image sensor, the pixel circuit comprising: a pinned photodiode (PD) coupled to a sense node (SN); a comparator (104) configured: to compare, during a first read phase, a first sense node voltage (Vpix) at the sense node (SN) with a first voltage ramp (V_RAMP) and to generate an output signal (WRITE) if the voltage of the first voltage ramp (V_RAMP) crosses the first sense node voltage (Vpix); and, otherwise, to compare, during a second read phase, a second sense node voltage (Vpix) at the sense node (SN) with a second voltage ramp (V_RAMP) and to generate an output signal (WRITE); and a memory (108) configured to receive a digital ramp (DATA_RAMP), and to store, in response to the output signal (WRITE) of the comparator (104), a value of the digital ramp (DATA_RAMP) to form pixel data (DATA_PIX).

The invention provides a dynamic target tracking method and device for an anti-drowning unmanned aerial vehicle and a storage medium, and the method comprises the steps: carrying out the preprocessing of water surface monitoring data collected by an RGB camera and an event camera which are carried on the unmanned aerial vehicle, and obtaining an event frame sequence and a water surface illumination quality evaluation result; performing water surface environment adaptability enhancement on the RGB image according to the event frame sequence and the water surface illumination quality evaluation result to obtain a time sequence registration relation and an enhanced RGB image sequence; performing underwater target feature fusion processing according to the enhanced RGB image sequence and the event frame sequence to obtain an underwater human body target feature template and an initial tracking position; and according to the underwater human body target feature template and the initial tracking position, carrying out continuous target tracking processing under a water surface shielding condition to obtain an underwater target tracking trajectory. According to the invention, the defect of the RGB camera in a complex water surface environment is compensated by using the high dynamic range characteristic of the event camera, and the target tracking precision and stability in an anti-drowning scene are improved.

The invention provides a cultural relic image acquisition system combining a depth-of-field stacking technology and a high dynamic range (HDR) splicing technology, and belongs to the technical field of image acquisition. According to the system, depth-of-field stacking and HDR splicing technologies are combined, focal plane positions and various exposure times are set firstly, and different exposure images are collected by each focal plane. And carrying out space alignment and brightness correction on the image by adopting an image registration algorithm based on feature points, and carrying out HDR synthesis on different exposure images of each focal plane to obtain a high dynamic range image. And calculating a definition evaluation value through a Laplace operator, and selecting the pixel with the highest definition to generate a panoramic depth image. Image feature point matching and alignment are carried out on multiple panoramic depth synthesis, seamless fusion is carried out through a smooth transition function based on weight, and the problems of splicing gaps and inconsistent brightness are eliminated. And finally, enhancing details by using a multi-scale Retinex algorithm, and performing color correction by using a standard color card to generate a high-precision and large-format cultural relic panoramic depth image.

The invention provides a self-adaptive reverse multiple exposure phase fusion technology, and relates to the field of high-dynamic-range three-dimensional reconstruction. The method includes the following steps that firstly, a saturated pixel mask matrix is obtained by judging pixels of a high-reflection area, then uniform grey-scale maps are selected step by step through a gradient adjustment method to obtain the optimal local projection intensity, and local coding and decoding are conducted on the high-reflection area through the optimal projection intensity pattern. And replacing an invalid phase value in the high-SNR image with a local phase value under the optimal projection intensity through an edge matching and phase fusion technology, and obtaining high-quality complete three-dimensional imaging information through coordinate mapping after a complete phase diagram is obtained. According to the method, the optimal projection intensity is adaptively determined through a small number of patterns, meanwhile, the high SNR of imaging patterns is guaranteed, and high-precision complete three-dimensional imaging is carried out on an HDR object. Experimental results show that in a traditional method, a large number of point clouds are lost in a saturated pixel area, the imaging integrity and the measurement precision cannot achieve the normal imaging effect, and the algorithm of the invention shows high integrity and high point cloud density in a dark area and a high-reflection area.

The invention relates to a semantic vector map robust construction method and device, a storage medium and a program product, which are applied to the technical field of automatic driving perception, and the method comprises the following steps: obtaining a multi-exposure original image; preprocessing the multi-exposure original image to obtain a high dynamic range image; bEV feature extraction and enhancement are carried out on the high dynamic range image to obtain a multi-scale BEV feature and an enhanced BEV feature; processing the enhanced BEV features to obtain a mixed query, and processing the mixed query and the multi-scale BEV features through a multi-layer mask activated instance decoder to obtain an embedded query and a coordinate query; processing the enhanced BEV features through a mask feature construction unit to obtain binary mask features, determining local block features based on coordinate query, and adjusting lane point coordinates based on the local block features, embedded query and a multi-layer perceptron to obtain a semantic vectorization map. The method has the effect of improving the accuracy and robustness of map construction.