A system includes a memory and at least one processor to acquire a hyperspectral image of a food object by an imaging device, the hyperspectral image of the food object comprising a three-dimensional set of images of the food object, each image in the set of images representing the food object in a wavelength range of the electromagnetic spectrum, normalize the hyperspectral image of the food object, select a region of interest in the hyperspectral image, the region of interest comprising a subset of at least one image in the set of images, extract spectral features from the region of interest in the hyperspectral image, and compare the spectral features from the region of interest with a plurality of images in a training set to determine particular characteristics of the food object and determine that the hyperspectral image indicates a foreign object.

The present disclosure accurately detects the position of a shelf label disposed on a display shelf. A shelf label detection device may be provided with a shelf label position correction unit and a shelf label position identification unit. The shelf label position correction unit corrects a manual shelf label position set including a shelf label position that has been set in advance for a reference camera image, using an automatic shelf label position set which includes a shelf label position detected from a monitoring camera image using image recognition, thereby generating a corrected shelf label position set. The shelf label position identification unit uses the corrected shelf label position set to identify the shelf label position in the monitoring camera image.

The feature extraction device according to one aspect of the present disclosure comprises: a reliability determination unit that determines a degree of reliability with respect to a second region, which is a region that has been extracted as a foreground region of an image and is within a first region that has been extracted from the image as a partial region containing a recognition subject, said degree of reliability indicating the likelihood of being the recognition subject; a feature determination unit that, on the basis of the degree of reliability, uses a first feature which is a feature extracted from the first region and a second feature which is a feature extracted from the second region to determine a feature of the recognition subject; and an output unit that outputs information indicating the determined feature of the recognition subject.

According to one embodiment, an image analysis device includes one or more processors configured to receive input of an image; calculate feature amount information indicating a feature of a region of the image; recognize a known object from the image on the basis of the feature amount information, the known object being registered in learning data of image recognition; recognize a generalization object from the image on the basis of the feature amount information, the generalization object being generalizable from the known object; and output output information on an object identified from the image as the known object or the generalization object.

A method for determining a property of an object is disclosed, which includes: recording a first image of the object from a first direction; recording a second image of the object from a second direction; determining a first position in the first image, the first position representing a location of the object, and a second position in the second image, the second position representing the same location of the object, for a multiplicity of locations of the object; and calculating a value of an object property for each of the multiplicity of locations of the object. The value assigned to a location of the multiplicity of locations of the object is calculated using an intensity value at the first position, which represents the location, in the first image and an intensity value at the second position, which represents the location, in the second image.

An object identification method determines whether a first monitoring image and a second monitoring image captured by a monitoring camera apparatus have the same object. The object identification method includes acquiring the first monitoring image at a first point of time to analyze a first object inside a first angle of view of the first monitoring image, acquiring the second monitoring image at a second point of the time different from the first point of time to analyze a second object inside the first angle of view of the second monitoring image, estimating a first similarity between the first object inside the first angle of view of the first monitoring image and the second object inside the first angle of view of the second monitoring image; and determining whether the first object and the second object are the same according to comparison result of the first similarity with a threshold.

In one embodiment, a method includes accessing a stream of F video frames, where each of the F video frames includes N patches that are non-overlapping, generating an initial embedding vector for each of the N×F patches in the F video frames, generating a classification embedding by processing the generated N×F initial embedding vectors using a self-attention-based machine-learning model that computes a temporal attention and a spatial attention for each of the N×F patches, and determining a class of the stream of video frames based on the generated classification embedding.

A method, apparatus, and device for labeling images of PCBs includes obtaining an image to be tested; comparing the image to be tested to a reference image to generate an image mask, the image mask includes several connected domains; detecting defects of the image to be tested; when at least one defect detected in the image to be tested, obtaining a coordinate of the at least one defect; based on a central coordinate of the connected domains and the coordinate of the at least one defect, determining the connected domains to be defect connected domains or normal connected domains; generating a first image mask and a second image mask; and processing the first image mask and the second image mask with the image to be tested to obtain a defect element image corresponding to the defect connected domains and a normal element image corresponding to the normal connected domains.

Apparatuses, systems, and techniques are presented to track objects represented in images or video data. In at least one embodiment, motion information corresponding to one or more objects within one or more digital images is used to determine a representative set of motion information corresponding to the motion information for the one or more objects.

This disclosure describes methods, non-transitory computer readable storage media, and systems that utilize image-guided model inversion of an image classifier with a discriminator. The disclosed systems utilize a neural network image classifier to encode features of an initial image and a target image. The disclosed system also reduces a feature distance between the features of the initial image and the features of the target image at a plurality of layers of the neural network image classifier by utilizing a feature distance regularizer. Additionally, the disclosed system reduces a patch difference between image patches of the initial image and image patches of the target image by utilizing a patch-based discriminator with a patch consistency regularizer. The disclosed system then generates a synthesized digital image based on the constrained feature set and constrained image patches of the initial image.