34 results about "Sky imaging" patented technology

A multifunctional sky camera system and techniques for the use thereof for total sky imaging and spectral irradiance/radiance measurement are provided. In one aspect, a sky camera system is provided. The sky camera system includes an objective lens having a field of view of greater than about 170 degrees; a spatial light modulator at an image plane of the objective lens, wherein the spatial light modulator is configured to attenuate light from objects in images captured by the objective lens; a semiconductor image sensor; and one or more relay lens configured to project the images from the spatial light modulator to the semiconductor image sensor. Techniques for use of the one or more of the sky camera systems for optical flow based cloud tracking and three-dimensional cloud analysis are also provided.

The invention provides a sun irradiance predication method based on full-sky imaging equipment and a device thereof. The method comprises the steps of firstly determining a sun irradiance predicated value based on a cloud cluster track, determining a sun irradiance predicated value based on weather forecasting mode and a sun irradiance actual measured value, and finally realizing final predicationof the sun irradiance. In the sun irradiance predication method based on the full-sky imaging equipment, effect of cloud to the sun irradiance is considered in determining a clear sky irradiance; effect of quick cloud change to the sun irradiance is considered in determining a cloud cluster movement track; and furthermore in a process of determining the sun irradiance predicated value based on the weather forecasting mode, physical processes such as weather process and cloud change are considered; comprehensive consideration is realized and furthermore accuracy of a sun irradiance predicatedresult is improved.

The invention relates to a cloud base height inversion method based on a foundation cloud picture. The cloud base height inversion method comprises the following steps: (1) selecting places, deployingtwo whole-sky imaging instruments, setting a field angle of the two whole-sky imaging instruments to be not greater than 130 degrees, enabling the two whole-sky imaging instruments to perform synchronous observation, and enabling the observed foundation cloud picture to have an overlapped area; (2) acquiring the foundation cloud picture synchronously observed by the two all-sky imaging instruments, performing the image matching for the two foundation cloud pictures, and selecting the overlapped area with the smallest difference; and (3) for a cloud cluster selected in the overlapped area in amatching manner, and a cloud base height is inversed by utilizing a geometric relation to obtain an inversion result. Compared with the satellite cloud picture observed from the cloud top in the existing inversion method, the foundation cloud picture is observed on the ground by the whole-sky imaging instruments, i.e., the foundation cloud picture is observed from the cloud base, so that the accuracy of the cloud base height inversion result is improved, and the effect is more significant for the thick cloud and the low cloud. In addition, the application range of the whole-sky imaging instrument is also expanded.

The invention discloses a large-aperture spliced telescope co-phase method, device and equipment and a storage medium, and the method comprises the steps: carrying out the star sky imaging through a large-view-field terminal, and obtaining first star point images of different view fields as a reference PSF image; translating the sub-mirror on the outermost ring of the spliced telescope to the maximum range to obtain second star point images of different fields of view as occlusion PSF images; performing Fourier transform on the reference PSF image and the shielded PSF image to obtain an OTF image difference, obtaining wrapping phases of different fields of view, and unwrapping the wrapping phases; and training a neural network by taking the phase information obtained after unwrapping and the conjugate graph thereof as inputs and taking the previous 37-order Zernike polynomial as an output, obtaining wavefront information of different fields of view through the trained neural network, and performing resolving to obtain the actual surface shape of the measured element. In this way, high-resolution wavefront detection data can be obtained by combining the OTF technology with the neural network, and the actual surface shape of the element can be accurately obtained.

The embodiment of the invention discloses an all-sky imager radiation calibration method based on a fixed star radiation spectrum and flux. The radiometric calibration method comprises the steps of selecting a target fixed star in an all-sky image, determining a zenith angle corresponding to the target fixed star at the moment of the all-sky image at an aurora observation station, and acquiring atmospheric transmissivity at the zenith angle corresponding to the target fixed star on an aurora observation spectral line according to the zenith angle; calculating to obtain photon radiation flux of the target fixed star on the aurora observation spectral line and a first solid angle corresponding to the target fixed star, and determining fixed star radiation brightness of the target fixed star according to the photon radiation flux and the first solid angle; and determining a detector output signal value corresponding to the target fixed star, and determining the radiation coefficient of the all-sky imager according to the detector output signal value, the atmospheric transmissivity and the fixed star radiation brightness. According to the technical scheme of the embodiment of the invention, radiometric calibration can be quickly and accurately carried out on the all-sky imager without geographical limitation, and the radiometric calibration of the all-sky imager is more flexible.

The present invention discloses a corrugated sensor, a corrugation reconstruction method and application thereof. The corrugated sensor comprises: a sampling array, for sampling the refracted light passing through the corrugation; a reflecting plate, wherein the reflecting plate comprises a light receiving surface, and the light receiving surface is configured to receive the sampled refracted light to form sampling spots; a first imaging element, for imaging the light receiving surface to obtain distribution information of the sampling spots; and a control element, for reconstructing the corrugation distribution according to the distribution information of the sampling spots, relative position information of the reflecting plate and the first imaging element, position information of the array unit in the sampling array corresponding to the sampling spot, and direction information of the incident light corresponding to the sampled refracted light. According to the technical scheme of the present invention, a reflecting plate is used as the spot imaging element for sampling the refracted light, so that the misalignment flare phenomenon of transmissive imaging is avoided, the depth applicable to the underwater is increased, and the imaging correction effect in the water surface-to-sky imaging application is improved.

The invention provides a satellite-to-ground camera imaging system for high-precision direction measurement of an optical remote sensing satellite, a field of view of a satellite-to-ground camera is divided into an over-the-sky field of view and an over-the-ground field of view, and the two fields of view are in a common-optical-path integrated design, so that a visual axis relation between an over-the-sky imaging module and an over-the-ground imaging module is stable; therefore, through common-path imaging on the sky and the ground and matching with other position resolving algorithms, high-precision calibration, positioning and even surveying and mapping tasks on the ground object position can be completed; based on the stable visual axis relation between a satellite-to-ground camera to the over-the-sky imaging module and the over-the-ground imaging module, the optical remote sensing satellite direction correction system firstly determines the visual axis direction of the to-the-ground imaging module in combination with the star map orientation acquired by the over-the-sky imaging module; and image matching is carried out according to image data obtained by ground imaging of the main camera and ground imaging data of the satellite-to-ground camera so as to further correct errors of the visual axis pointing of the main camera, so that the visual axis pointing of the main camera is calculated with high precision, and the surveying and mapping precision of the main camera is improved.

The invention discloses a short-time cloud picture tracking method based on an optical flow method. The method comprises: firstly, collecting real-time sky pictures through an all-sky imager, and removing irrelevant factors such as buildings in the images are removed; secondly, utilizing RBG color information of the image to separate pixels of the cloud from pixels of the sky; and finally, comparing the same cloud pixels of the two frames to obtain an edge position and a speed vector of the cloud, predicting the edge position of the cloud at the next moment, and comparing the edge position with the sun position to realize prediction. The method can achieve the recognition and prediction of the position of the cloud picture, is higher in accuracy, is high in speed, is high in universality,is good in robustness, and can be used for an irradiation intensity prediction or solar photovoltaic power output prediction system.

An autonomous system for geographic locating may include a camera unit capable of capturing a present time image of a night sky, a processor operable to execute instructions accessible in memory, the processor capable of implementing a machine learning positioning algorithm comprising a finite sequence of instructions, the machine learning positioning algorithm comprising a training module operable in a training mode with a training dataset to train same, the machine learning positioning algorithm including a prediction module operable in a prediction mode with a live dataset to provide a prediction of an inferred geographic location of capturing the present time image.

The invention relates to an aurora monitoring system. The aurora monitoring system comprises a fish-eye lens for aurora full-sky imaging, a low-light industrial camera for converting the aurora full-sky imaging into an aurora analog video signal, a camera controller for controlling the low-light industrial camera, and a video encoder for converting the aurora analog video signal into a digital video and providing the outside with an online video streaming function by connecting the digital video with a network, wherein the fish-eye lens is installed at the front end of the low-light industrial camera, the low-light industrial camera is connected with the video encoder through a bayonet nut connector (BNC) cable, and the camera controller is electrically connected with the low-light industrial camera. According to the aurora monitoring system, good aurora imaging is achieved under an extreme weak light condition, the online video streaming function can be provided, automatic unattended observation is achieved, and the aurora monitoring system has the advantages of being convenient to install and operate, low in cost, good in imaging quality, and capable of performing on-line video streaming output, timing image capture, and meeting requirements of aurora activity real-time monitoring.

The embodiment of the invention discloses a geometric calibration method of an all-sky imager based on a gradient descent method. The geometric calibration method comprises the following steps: selecting a target star point in an all-sky image, and determining star point position coordinates of the target star point in an all-sky image coordinate system and a real azimuth angle and a real zenith angle of the target star point at an aurora observation station at the moment of the all-sky image; constructing an error function based on the real azimuth angle and the zenith position and the geographic orientation in the all-sky image, and determining a real zenith position and a real geographic orientation according to the error function; and determining the imaging radius of the target star point according to the star point position coordinates and the real zenith position, and obtaining the relation between the imaging radius and the real zenith angle in the all-sky image through fitting according to the real zenith angle and the imaging radius. According to the technical scheme of the embodiment of the invention, geometric calibration can be quickly and accurately carried out on the all-sky imager without being limited by regions, and the geometric calibration of the all-sky imager is more flexible.

The invention provides a solar energy resource ultra-short time forecasting method based on ground-based cloud pictures. The solar energy resource ultra-short time forecasting method comprises the following steps: determining historical cloud information based on a plurality of original ground-based cloud pictures obtained by an all-sky imager in a past preset time period; establishing a cloud cover forecasting model according to the historical cloud information; determining a forecast cloud cover according to the cloud cover forecasting model, wherein the forecast cloud cover is the cloud cover in a future preset time period; establishing a clear sky radiation forecast model according to historical clear sky radiation data in a past preset time period; establishing a cloud attenuation model according to the historical cloud information; determining a cloud sky radiation model according to the clear sky radiation forecast model and th'e cloud attenuation model; and determining solar radiation output in the future predetermined time period by using the cloud sky radiation model according to the forecast cloud amount.

The invention relates to the field of meteorological instrument calibration, in particular to an all-sky imager calibration method based on lunar orientation, which comprises the following steps: S1, fixing and installing an all-sky imager, and recording the longitude and latitude of the installation position of the all-sky imager; s2, setting the shooting interval of the all-sky imager, automatically collecting and obtaining a series of first foundation cloud pictures, and recording the shooting date and shooting time of each first foundation cloud picture; s3, carrying out image recognition to obtain a centroid coordinate of the moon in the first foundation cloud picture; s4, calculating the lunar elevation angle H and the lunar azimuth angle L at each shooting time according to the shooting date, the shooting time and the longitude and latitude of the installation position; and S5, performing parameter fitting on the all-sky imager according to the coordinates of the centroid of the moon, the elevation angle of the moon and the azimuth angle of the moon so as to complete calibration of the all-sky imager. The method has the technical characteristics of simplicity and rapidness in calibration, high calibration precision and low calibration cost.

The invention relates to a regional power grid load prediction method and device, which belong to the technical field of electric power, and solve the problem that data information contained in unstructured meteorological factors is lost when modeling is carried out through structured meteorological data, so that the accuracy of load prediction is influenced. The method comprises the steps of determining meteorological data influencing loads in all meteorological partitions in a regional power grid, wherein the meteorological data comprise air pressure, temperature, precipitation, relative humidity, wind speed, wind direction, date types and cloud picture data obtained through shooting of an all-sky imager, preprocessing the meteorological data, establishing a cloud picture classification and discrimination model of a Gabor filter-convolutional neural network, and performing prediction and classification processing on the preprocessed cloud picture data by using the discrimination model, fusing the classified cloud picture data with other meteorological data to form a meteorological data set, establishing a load prediction model, and predicting the load of each meteorological partition by using the load prediction model. And the accuracy and precision of load prediction are improved.

The invention discloses a method and system for predicting the irradiance of the ground surface through a sky image-irradiance model. The method proposed by the invention includes the following steps: establishing a prediction model of the irradiance of the ground surface, and using historical sky images to extract features to obtain the The surface irradiance prediction model is trained; a plurality of current sky images separated by a first preset time period are obtained; the current sky image is corrected to obtain a corrected current sky image; cloud displacement vector calculation method is used to obtain Cloud cluster displacement vector between adjacent sky images; Obtain the cloud cluster information that will block the sun part within the second preset time end for each current sky image; Obtain the first average RGB value and the first distance value; The mirror surface used by the sky imager is a fisheye convex mirror, so there is a certain distortion in the captured sky image; after correction processing, the accuracy of the first average RGB value and the first distance value can be improved.

The invention relates to a multi-view-field starry sky observation satellite target attitude planning method, which comprises the following steps of: 1, acquiring seasons or months of starry sky observation of a satellite, a satellite orbit height and a stray light suppression angle of imaging optical equipment; 2, according to the projection relation between the selectable view field and the observable sky area of the optical equipment, providing the selectable sky area of each piece of optical equipment; 3, selecting an observation point set according to the distribution of fixed stars in the selectable sky area, obtaining the direction of each optical device under the observation point under the celestial coordinate system, judging whether the optical device is in the available sky area, and deleting unavailable observation points; 4, counting the number of fixed stars in the field of view of the camera under the available observation points, and optimizing the available observation points; and 5, repeating the steps 3 and 4 until a satisfactory observation point set is found, and outputting the right ascension and declination of the optimal observation point and a satellite attitude conversion matrix. According to the invention, ideal observation sky areas and target attitudes are provided for multi-view-field star sky imaging of remote sensing satellites and astronomical observation satellites.