Abstract: Systems and methods for altering colors of pixels of images produced by cameras using a lens. A plurality of generated images are received, and color channel values and luminance values of image pixels are determined. Benchmark U-color thresholds and V-color thresholds are established based on the color channels and luminance associated with the pixels. A first image generated using a first lens of a first camera is received. U-color values and V-color values associated with first pixels of the first image are determined. A uniformity score associated with the first lens is received. The benchmark U-color and V-color thresholds are altered based on the uniformity score. The U-color values and V-color values of the first pixels are corrected in response to the U-color values being within the altered U-color thresholds and the V-color values being within the altered V-color thresholds.

Abstract: Implementations are disclosed for analyzing extracted individual plant components to make agronomic inferences about entire composite plant organs from which the individual plant components were harvested, and for using those agronomic inferences for various purposes. In various implementations, individual plant component(s) may be sampled from multiple plant components removed from composite plant organ(s) that previously included the plant components. Digital image(s) may be captured of the sampled individual plant component(s) and processed based on machine learning model(s) to generate agronomic inference(s) about the composite plant organ(s) that previously included the plurality of plant components. Based on the agronomic inference(s), computing device(s) may render output that includes a diagnosis or recommendation for the grower about the field or the crops, and/or agricultural equipment (e.g., robots) may be operated automatically.

Abstract: Techniques for identifying a geographic location of a wind turbine using an image. An image depicting a geographic area is received. A location is identified, in the image, relating to a place where a wind turbine tower depicted in the image meets ground depicted in the image. This includes identifying two objects relating to the wind turbine tower in the image, and distinguishing a bottom end of the wind turbine tower from a top end of the wind turbine tower based on the identified two objects. A geographic location corresponding to the determined image location is determined, based on the image and the determined image location. A location of the wind turbine tower, in a database of wind turbine locations, is updated based on the geographic location.

Abstract: An image processing method includes computing a blur kernel of an input image using a kernel estimation model, performing kernel-based deblurring on the input image using the blur kernel to obtain a deconvolved image, and generating an output image by performing kernel-free deblurring based on the deconvolved image.

Abstract: Systems and methods for performing one or more medical imaging analysis tasks are provided. A sequence of medical images is received. One or more patches are extracted from each image of the sequence of medical images. Spatio-temporal features are extracted from the one or more extracted patches using a machine learning based encoder network. One or more medical imaging analysis tasks are performed based on the extracted spatio-temporal features. Results of the one or more medical imaging analysis tasks are output. The machine learning based encoder network is trained by receiving a sequence of training medical images. Patches of a first set of images of the sequence of training medical images are masked according to a first masking strategy. Patches of a second set of images of the sequence of training medical images are masked according to a second masking strategy.

Abstract: The present invention provides a method for estimating hydrocarbon saturation of a hydrocarbon-bearing rock from a resistivity log and a rock image. The image is segmented to represent either a pore space or solid material in the rock. An image porosity is estimated from the segmented image, and a corrected porosity is determined to account for the sub-resolution porosity missing in the image of the rock. A corrected cementation exponent of the rock is determined from the image porosity and the corrected porosity and is used to estimate the hydrocarbon saturation. A backpropagation-enabled trained model can be used to segment the image. A backpropagation-enabled method can be used to estimate the hydrocarbon saturation using an image selected from a series of 2D projection images, 3D reconstructed images and combinations thereof.

Abstract: Methods and systems are provided for use in evaluating quality of pollen grains. One example method includes germinating pollen grains of a crop in a fluid media and inserting a sample of the germinated pollen grains and fluid media into a flow chamber of a fluid imaging system. The sample flows through the flow chamber at a predefined rate, and images of the sample are captured by a camera of the fluid imaging system. The method then includes comparing the captured images of the sample to reference images specific to the crop and, based on the comparison, generating a pollen germination score for the sample, the pollen germination score indicative of germinated pollen grains in the sample.

Abstract: Apparatuses, systems, and techniques to determined a pose of an object from a plurality of images. In at least one embodiment, the pose of an object is determined from at least two images of a video sequence using one or more neural networks, in which the neural network produces a distribution of pose information that is filtered to determine the current pose.

Abstract: Aspects relate to techniques for estimating relative pose leveraging lane markers. A first wireless communication device may receive a message from a second wireless communication device including a plurality of visual features (e.g., keypoints) of an image captured by the second wireless communication device, lane features identifying two (or more) traffic lane markers, and respective two-point two-line (2L2P) metrics for pairs of keypoints based on the traffic lane markers. The first wireless communication device may then obtain additional features from an additional image captured by the first wireless communication device and identify matching keypoints of the additional features for which the respective 2L2P metrics are satisfied. The first wireless communication device may then calculate a relative pose of the first wireless communication device with respect to the second wireless communication device based on the matching features.

Abstract: A LiDAR system includes alight emitter, a light detector, and a controller. The controller is programmed to: activate the light emitter to emit a series of shots into a field of view of the light detector; activate the light detector to detect shots reflected from an object in the field of view; record the detected shots from a first subset of the series of shots; group a second subset of the series of shots not in the first subset; and based on the detected shots from the second subset of the series of shots, identify an object moving in the field of view.

Abstract: These disclosures relate to a method for detecting and removing personnel interference while measuring volume of an object. An image of an object is obtained, a background information of the image is removed, and the image is converted into a binary image. An outline of the object is detected from the binary image. A determination is made whether the outline extends towards a boundary of the field of view. If the outline of the object does not extend to the boundary of the field of view, the volume of the object is measured. Otherwise, an extending part is removed from the outline using a morphological method, if removable. If the extending part is removed, the volume of the object is measured based on an outline of the object obtained after the extended part is removed, and a class 1 warning is issued. Otherwise, a class 2 warning is issued and the measurement is ended. This approach improves efficiency of the object volume measurement.

Abstract: A process for aligning at least two different digital scan data each representing at least a part of a dentition of a patient includes scanning the patient's dentition so as to record volumetric 3-dimensional digital scan data in a volumetric coordinate system and scanning the patient's dentition so as to record topographic 3-dimensional digital scan data in a topographic coordinate system. The teeth or jaws of the patient's dentition are segmented and approximated from the digital scan data by grouping the digital scan data with respect to structures and approximating a surface of the segmented teeth using triangle meshes. A key point determination is performed for each tooth based on the triangle mesh representation to define 3-dimensional coordinates of a tooth surface feature in the volumetric or topographic coordinate system. The key points are aligned to obtain aligned digital scan data by coordinate transformation of one coordinate system.

Abstract: The present disclosure provides systems and methods for material decomposition. The systems may obtain scan projection data of a target object. The systems may determine corrected projection data by correcting, based on one or more pixel parameters, the scan projection data. The systems may also determine a reconstructed image by performing, based on the corrected projection data, image reconstruction. The systems may further determine density distribution images of at least two target materials of the target object by decomposing the reconstructed image.

Abstract: Disclosed in an embodiment according to the present disclosure is an electronic device comprising: a camera, a display, a sensor configured to acquire depth information about at least one object, and at least one processor electrically connected to the camera, the display and the sensor, wherein the at least one processor is configured to display, as a preview image on the display, an image acquired by driving the camera, acquire an input for designating a start point and an end point on a subject included in the preview image, acquire depth information about the subject through the sensor, correct the start point and the end point on the basis of the acquired depth information, and measure the length of the subject on the basis of the distance between the corrected start point and the corrected end point.

Abstract: Systems and methods for image reconstruction for parallel MR imaging are disclosed that receive a k-space single-coil measurement dataset that includes at least two k-space single-coil measurement sets, transforming the k-space single-coil measurement dataset to an estimated CSM using a coil sensitivity estimation module, and transforming the k-space single-coil measurement dataset and the estimated CSM into a final MR image using an MRI reconstruction module. In some aspects, the coil sensitivity estimation module and MRI reconstruction module include deep learning neural networks trained without the use of ground truth data.

Abstract: An infrared temperature measurement method, apparatus, device, and storage medium, which relate to the field of image processing is disclosed. The method comprises: determining a measurement temperature of a to-be-measured target based on a grayscale image captured to obtain a first measurement temperature; determining key point coordinates of the target based on a target image; determining a first distance based on the key point coordinates and configuration information of an infrared temperature measurement device, the first distance is a distance between the target and the infrared temperature measurement device; determining an actual temperature of the target based on the first measurement temperature and the first distance.

Abstract: Retrographic sensors described herein may provide smaller sensors capable of high-resolution three-dimensional reconstruction of an object in contact with the sensor. Such sensors may be used by robots for work in narrow environments, fine manipulation tasks, and other applications. To provide a smaller sensor, a reduced number of light sources may be provided in the sensor in some embodiments. For example, three light sources, two light sources, or one light source, may be used in some sensors. When fewer light sources are provided, full color gradient information may not be provided. Instead, the missing gradients in one direction or other information related to a three-dimensional object in contact with the sensor may be determined using gradients in a different direction that were provided by the real data. This may be done using a trained statistical model, such as a neural network, in some embodiments.

Abstract: A system and method and for searching for an element in and replacement a portion of a video is conducted by retrieving the video and generating a video index for the video, the video index indexing a plurality of elements of the video such as text, audio and video. To search in the vide, a search query is received that includes at least one of a text element, an audio element, an image element, or a video element. Features and embeddings are extracted from the search query before comparing the search query features with the video index to identify a portion of the video that includes one of the text element, audio element, image element or video element of the search query. The identified portion is then automatically replaced with a replacement portion to generate a revised video.

Abstract: An image processing apparatus divides scanned image data including page images obtained by scanning a plurality of documents for each page into image data of each document. The apparatus generates text data by performing character recognition processing for the plurality of page images, sequentially obtains a pair of page images in succession from the plurality of page images and then determines a document delimitation position based on text data of the two page images constituting the pair, and divides the scanned image data at the determined delimitation position. A vector corresponding to tokens is obtained by decomposing the text of each of the two page images constituting the pair is generated and input to a neural network model, the delimitation position is determined by using a score output from the neural network model and represents a possibility value that the two page images constituting the pair belong to different documents.

Abstract: Provided are a method and an apparatus for measuring an object quantity using a 2D image. A method for measuring an object quantity using a 2D image according to one embodiment of the present disclosure comprises learning an object region model in a first object image which is a pre-learning target, learning an object quantity model for a first object region in the first object image using a feature map extracted from the first object image, and measuring an object quantity of at least one of a second object region and a background region in a second object image which is a measurement target using the learned object region model and object quantity model.