A height estimation apparatus (10) according to the present disclosure includes an acquisition unit (11) for acquiring a two-dimensional image obtained by capturing an animal, a detection unit (12) for detecting a two-dimensional skeletal structure of the animal based on the two-dimensional image acquired by the acquisition unit (11), and an estimation unit (13) for estimating a height of the animal in a three-dimensional real world based on the two-dimensional skeletal structure detected by the detection unit (12) and an imaging parameter of the two-dimensional image acquired by the acquisition unit (11).

A method can include receiving (1) images of at least one subject and (2) at least one total mass value for the at least one subject. The method can further include executing a first machine learning model to identify joints of the at least one subject. The method can further include executing a second machine learning model to determine limbs of the at least one subject based on the joints and the images. The method can further include generating three-dimensional (3D) representations of a skeleton based on the joints and the limbs. The method can further include determining a torque value for each limb, based on at least one of a mass value and a linear acceleration value, or a torque inertia and an angular acceleration value. The method can further include generating a risk assessment report based on at least one torque value being above a predetermined threshold.

Embodiments of the disclosure provide methods and systems for representation learning from a biomedical image with a sparse convolution. The exemplary system may include a communication interface configured to receive the biomedical image acquired by an image acquisition device. The system may further include at least one processor, configured to extract a structure of interest from the biomedical image. The at least one processor is also configured to generate sparse data representing the structure of interest and input features corresponding to the sparse data. The at least one processor is further configured to apply a sparse-convolution-based model to the biomedical image, the sparse data, and the input features to generate a biomedical processing result for the biomedical image. The sparse-convolution-based model performs one or more neural network operations including the sparse convolution on the sparse data and the input features.

Systems and methods for utilizing Augmented Reality (AR) processes to track cables among a tangled bundle of cables are provided. An AR method, according to one implementation, includes a step of obtaining an initial captured image showing a bundle of cables. The AR method also includes the step of processing the initial captured image to distinguish a selected cable from other cables of the bundle of cables. Also, the AR method includes displaying the initial captured image on a display screen while visually augmenting an image of the selected cable to highlight the selected cable with respect to the other cables.

A method includes calculating a first medial-axis tree graph of a volume of an organ of a patient in a first computerized anatomical map of the volume, acquired at a first time. A second medial-axis tree graph is calculated, of a volume of the organ of the patient in a second computerized anatomical map of the volume, acquired at a second time that is different from the first time. A deviation is detected and estimated, between the first and second tree-graphs. Using the estimated deviation, the first and second medial-axis tree graphs are registered with one another. Using the registered first and second tree graphs, the first and second computerized anatomical maps are combined.

The present invention provides a system for identifying individual crowns of individual fruit trees using aerial images. A crown identification device 40 of the present invention includes an identification criterion determination unit 41 and a crown identification unit 42. The identification criterion determination unit 41 includes a first image acquisition section 411 that acquires a first aerial image including a plurality of individual fruit trees in a deciduous period in a fruit farm field, a skeleton extraction section 412 that processes the first aerial image to extract a whole crown skeleton including the plurality of individual fruit trees, a vertex extraction unit 413 that extracts vertexes of each crown skeleton corresponding to each individual fruit tree, and an identification criterion extraction section 414 that extracts a crown candidate region of a minimum polygonal shape including all the vertexes as an identification criterion for each individual fruit tree and extracts a centroid of the crown candidate region. The crown identification unit 42 includes a second image acquisition section 421 that acquires a second aerial image of the fruit tree farm field at the time of identifying a crown at the same scale as the first aerial image, a whole crown extraction section 422 that processes the second aerial image to extract a whole crown image including the plurality of individual fruit trees, and a crown identification section 423 that collates the crown candidate region and the centroid of the identification criterion with the whole crown image to identify a crown region of each individual fruit tree in the second aerial image.

An image processing system (10) according to an example aspect of the present disclosure includes: an image acquisition unit (102) configured to acquire a monochrome image; a pixel value correction unit (120) configured to correct a pixel value of the monochrome image based on information related to a pixel value; and a colorization generation unit (130) configured to generate a colorized image corresponding to the monochrome image from the corrected monochrome image by using a colorization prediction model trained by machine learning. With the present disclosure, it is possible to improve the reproduction accuracy of a color in colorization of a monochrome image.

The present disclosure provides a method and apparatus for an automatic parking system. The method includes: obtaining a parking-line segmentation image through a segmentation neural network; preprocessing the parking-line segmentation image to obtain at least one parking-line skeleton image; and calculating straight-line equations and straight-line endpoints for parking lines in the parking-line segmentation image based on the at least.

A method and a device for the characterization of living specimens from a distance are disclosed. The method comprises: acquiring an image of a living specimen and segmenting the image, providing a segmented image; measuring a distance to several parts of said image, providing several distance measurements, and selecting a subset of those contained in the segmented image. The method also processes the segmented image and the distance measurements referred to different positions contained within the segmented image by characterizing the shape and the depth of the living specimen and by comparing a shape analysis map and a depth profile analysis map. If a result of the comparison is comprised inside a given range, parameters of the living specimen are further determined including posture parameters, location or correction of anatomical reference points and/or body size parameters, and/or a body map of the living specimen is represented.

In an image having a blank region, the present invention estimates skeleton information having a part of a joint in the blank region. An estimation device (1) is provided with: an input part (130) for acquiring a first image that includes a first joint and does not include a second joint of the subject; a blank region expansion part (101) for generating a second image from the first image by expanding the blank region; and an estimation part (12) for estimating, using the second image and a prelearned estimation model, skeleton information that includes the joint position of the second joint located in the blank region.