15 results about "Parallel polarization" patented technology

The invention belongs to the technical field of polarization imaging, and particularly relates to a polarization dimension reduction imaging method, system and device based on principal component analysis, and the method comprises the steps: splitting a 0-degree polarization image from an original polarization image as a parallel polarization state image, and splitting a 90-degree polarization image as a vertical polarization state image; respectively extracting a background area and a target area; estimating the background intensity at the infinite distance in the original polarization image and the polarization degree of back scattering light; performing principal component analysis on the data matrix, and performing PCA dimension reduction and feature extraction to obtain a first feature value and a second feature value; and taking the first characteristic value as the total light intensity of the original polarization image, taking the second characteristic value as the light intensity difference between the parallel polarization state image and the vertical polarization state image, calculating the transmissivity of the original polarization image, and correcting the inherent radiation of the target area to obtain a final imaging image. According to the invention, the problem of light intensity estimation deviation at infinity in polarization imaging and the problem of transmissivity distortion caused by data redundancy are solved.

A dynamic 3D light scattering particle size distribution measurement device includes a polarization splitter for generating an s laser beam with perpendicular polarization and a p laser beam with parallel polarization that travels parallel to and spaced apart from the s laser beam. A lens is used to focus the s laser beam and the p laser beam onto a common focus area inside a sample holder holding a sample whose particle size is to be measured. An s light intensity detector measures a time-resolved s light intensity from s backscattered light from the focus area. An s-polarisation filter allows perpendicularly polarized light to pass, resulting in s measuring light, and includes a light detector for time-resolved intensity measurements of the s measuring light. A p light intensity detector measures a time-resolved p light intensity detects p backscattered light from the focus area, and includes a p polarization filter that allows parallel polarized light to pass. A light detector provides for time-resolved intensity measurement of the p measuring light, and an evaluation unit automatically calculates the particle size distribution from the time-resolved p light intensity and the time-resolved s light intensity.

The invention provides a flare-removing water surface target segmentation method and system based on polarization feature fusion, and the method comprises the steps: firstly constructing a polarization multi-feature deep fusion frame, and remarkably improving the recognition capability of a small target in a complex environment, such as strong dazzle light, through the combined utilization of the polarization features of the degree of polarization (DOLP), the polarization angle (AOP) and the parallel polarization radiation (PPR); secondly, designing a GFNet enhanced network based on a Retinex theory, carrying out coding separation and fusion on PPR and AOP images, inhibiting dazzle light, enhancing target details, and outputting a high-quality enhanced image (PPRen); and finally, constructing a CDDFuse fusion network and designing a two-stage training mechanism, firstly training the GFNet to enhance the network and optimize the dazzle light removal capability, and then training the CDDFuse fusion network by adopting PPRen and DOLP to realize frequency domain feature decomposition and fusion, thereby improving the target and background distinguishing capability and remarkably improving the segmentation precision and robustness of the system.

The invention relates to the technical field of disaster early warning, in particular to a target identification method based on laser radar multi-source data fusion, and the method comprises the steps: S1, transmitting a polarization laser beam to a monitoring area through a laser radar, and receiving a backscattering signal; s2, calculating a depolarization ratio of the back scattering signal, and distinguishing smoke particles and non-smoke particles based on the depolarization ratio; the depolarization ratio is the ratio of the intensity of a vertical polarization component to the intensity of a parallel polarization component in the back scattering signal; and S3, based on the distinguished smoke particle information, determining whether a fire danger occurs and triggering an alarm. According to the invention, by using the physical characteristic of the depolarization ratio, fire smoke and water vapor fog drops can be fundamentally distinguished, the false alarm rate in hazy weather is significantly reduced, the accuracy and reliability of fire danger identification are improved, and accurate identification and accurate positioning of fire danger smoke are realized by using the multi-dimensional detection capability of the laser radar.

The invention provides a variable temperature stress detection method based on a transverse wave birefringence sound-to-time ratio. Comprising the following steps: 1, calibrating a sample, and measuring propagation time in a parallel polarization direction and transverse wave propagation time in a vertical polarization direction; 2, calculating a sound-time ratio according to the obtained propagation time, and obtaining a stress calculation formula; 3, measuring transverse wave velocities in the sample at different temperatures to obtain a temperature-wave velocity coefficient, and measuring transverse wave propagation velocities in a parallel polarization direction and a vertical polarization direction in the sample at a calibration temperature to obtain temperature correction coefficients in the two directions; 4, correcting the propagation time of the transverse wave in the parallel polarization direction and the propagation time of the transverse wave in the vertical polarization direction by using the temperature correction coefficient obtained in the step 3; 5, substituting the corrected propagation time and the temperature correction coefficient into the stress calculation formula in the step 2 to obtain a shear wave birefringence temperature correction stress calculation formula based on the sound-time ratio; and 6, measuring the propagation time and the temperature difference of the transverse wave in the sample, and substituting into a correction stress calculation formula to obtain a corresponding stress value.

The invention discloses a giant clam shell waste image data acquisition processing method and system, and relates to the technical field of giant clam shell waste image processing, and the method comprises the steps: obtaining a black level map, a response uniformity correction map and two working wavelength channel gains based on a black field map and a white board map; the method comprises the following steps: collecting four images under parallel polarization and orthogonal polarization, correcting, uniformizing, calculating a common mode item and a difference item, and normalizing to obtain a dual-polarization dual-wavelength interference difference index graph; and generating a confidence map in combination with the saturated pixel mask and local noise estimation, and performing robustness to obtain an effective pixel mask and a statistic set, and finally forming a sample data packet. According to the invention, anti-interference, stable, comparable and traceable detection of the tridacna waste image data can be realized.

The application provides a SERF inertial measurement device nuclear spin lossless flip method, relates to the technical field of quantum precision measurement and inertial navigation, and lossless inverts the nuclear spin polarization direction by applying a specific radio frequency magnetic field to the SERF inertial measurement device. The frequency and amplitude of the applied radio frequency magnetic field change over time, so that the effective magnetic field experienced by the nuclear spin in the rotating coordinate system changes slowly from initially parallel to the polarization direction to the opposite polarization direction, thereby realizing lossless inversion of the nuclear spin polarization direction. The application has important significance for realizing online rapid measurement of nuclear polarization rate, developing a new type of inertial measurement output scheme and further suppressing system slow drift error.

The invention relates to the technical field of data analysis, in particular to a method and device for distinguishing old and new paper sheet articles and a medium, and the method comprises the steps: obtaining a collection image of a target paper sheet article based on a camera system; performing multi-scale linear feature detection on the acquired image to obtain a crease loss feature value; determining a stain loss characteristic value and a texture degradation characteristic value according to the acquired image; and according to the crease loss characteristic value, the stain loss characteristic value and the texture degradation characteristic value, determining a new and old evaluation index of the target paper sheet article, and determining the new and old evaluation index of the target paper sheet article according to a parallel polarization image and a vertical polarization image collected by a dual-polarization camera system. And reliable data support is provided for new and old evaluation of the paper sheet-shaped articles.

The invention provides a nuclear spin lossless inversion method for an SERF inertial measurement device, relates to the technical field of quantum precision measurement and inertial navigation, and aims to perform lossless inversion on a nuclear spin polarization direction by applying a specific radio frequency magnetic field to the SERF inertial measurement device. The frequency and amplitude of the applied radio frequency magnetic field change along with time, so that an effective magnetic field sensed by nuclear spin in a rotating coordinate system slowly changes from an initial parallel polarization direction to an anti-parallel polarization direction, thereby realizing lossless inversion of the nuclear spin polarization direction. The method provided by the invention is of great significance in realizing on-line rapid measurement of the nucleon polarizability, developing a novel inertial measurement output scheme and further suppressing the slow drift error of the system.

The invention discloses a radar vertical detection sand and dust weather monitoring method, system and device and a medium, and belongs to the technical field of sand and dust weather monitoring, and the method comprises the steps: collecting a laser back scattering echo signal; the method comprises the following steps: carrying out polarization separation processing on a laser back scattering echo signal to obtain a parallel polarization signal component and a vertical polarization signal component; calculating a signal intensity ratio through the parallel polarization signal component and the vertical polarization signal component to obtain a depolarization ratio parameter; comparing and judging the depolarization ratio parameter with a preset sand and dust threshold value, and obtaining sand and dust particle identification information according to a judgment result; performing spatial height analysis and vertical height statistics through the sand and dust particle identification information to obtain a sand and dust weather detection result and sand and dust vertical profile data; and comprehensively outputting the sand and dust weather detection result and the sand and dust vertical profile data as sand and dust weather monitoring information. According to the invention, a laser radar backscattering polarization separation technology is adopted, and accurate identification of sand and dust particles is realized by analyzing a depolarization ratio parameter.

The invention discloses a fruit and vegetable quality grading processing method and system based on machine vision, and relates to the technical field of fruit and vegetable quality grading processing, and the method comprises the steps: collecting parallel polarization component images and vertical polarization component images of fruits and vegetables at different incident angles; performing brightness profile extraction on the parallel polarization component image and calculating millimeter half-peak width; carrying out equivalent motion broadening compensation in combination with the conveying speed and the exposure time to obtain a corrected half-peak width, and calculating a polarization retention rate based on a gray scale ratio of the parallel polarization image to the vertical polarization image; generating a coupling index by using the corrected half-peak width and the polarization retention rate, constructing a sliding reference set based on historical data, and determining upper and lower bound thresholds through a robust statistical method; and finally, grading the quality of the fruits and vegetables according to the threshold value, and dynamically updating the sliding reference set. The fruit and vegetable quality grading stability and accuracy can be improved.

This invention relates to the field of jewelry testing technology, and provides a light source adaptive adjustment device for jewelry testing, including a base and a worktable. A multispectral array light source is arranged above the worktable, and a camera is arranged on one side of the multispectral array light source. The multispectral array light source provides ultraviolet light, white light, near-infrared light, and various monochromatic visible light. A lens switching mechanism is installed at the output end of the multispectral array light source, which can switch the lens to a condenser lens or a diffuser lens. A third motor is fixedly installed on the camera, and a polarization mode switching disk is installed on the output shaft of the third motor. The polarization mode switching disk consists of four sets of polarization mode plates, which correspond to no polarization mode, single polarization mode, cross polarization mode, and parallel polarization mode, respectively. Through the coordinated automatic switching of the lens and polarization mode, multi-level and multi-dimensional observation from the surface to the interior is realized, improving the testing efficiency.

The application discloses a living body recognition device, which comprises a light source for projecting infrared light at a first power; a parallel polarizer located on the light path of the infrared light to irradiate the target object with the polarized state; a multifunctional polarizer located on the reflection signal path of the target object and comprising at least two states of parallel polarization and vertical polarization; an infrared sensor for receiving the reflection signal passing through the multifunctional polarizer; and a processor for calculating a first AOP according to the reflection signals of the parallel polarization and the vertical polarization, and determining whether the target object is a living body. The application has the characteristics of multi-dimensional information, good noise suppression, high accuracy and good recognition effect.

The invention discloses a living body skin detection method based on multispectral depolarization. The method comprises the steps that optical parameters are initialized and matched with the environment; starting a hardware system, synchronously acquiring facial RGB video data of the subject in parallel and cross polarization modes, and covering multiple parts such as cheeks and necks; key frame images in parallel polarization and cross polarization modes are extracted from the collected video, a B channel interfered by mirror reflection is abandoned, and R and G channel data are reserved; the extracted dual-polarization image data is preprocessed, noise is suppressed, and the subsequent calculation efficiency is improved; the R / G channel ratio in the dual-polarization mode is calculated, the intensity difference is obtained after normalization processing, and a living body skin heat map is constructed; the initial heat map H is optimized, invalid areas are removed, features are smoothed, living body skin area features are strengthened, and meanwhile visualization is completed; and judging whether the living body skin exists in the image or not according to the feature signal of the living body skin area in the heat map, and outputting a detection result to complete living body recognition. According to the method, image skin area detection and recognition are carried out through dual-polarized light source time-sharing irradiation and a lightweight algorithm, and meanwhile high recognition precision, high robustness and real-time performance are taken into consideration.