A system and a method for searching an image within another image are disclosed. The method includes producing template edge images and target edge images, having image scales, based on determination of edge gradients of a template image and a target image in one or more directions. The template image indicates an image to be searched. The target image indicates another image within which the image needs to be searched. Further, images comprising correlation coefficient values are produced for each of the directions by computing correlation coefficients between the template edge images and the target edge images. At least one local peak is identified from each of the images comprising the correlation coefficient values. Spatial locations along with the correlation coefficients corresponding to the local peak are determined. Thereafter, a presence of the template image in the target image is identified based upon an intersection of the spatial locations.

Methods and image processing systems are provided for determining a dominant gradient orientation for a target region within an image. A plurality of gradient samples are determined for the target region, wherein each of the gradient samples represents a variation in pixel values within the target region. The gradient samples are converted into double-angle gradient vectors, and the double-angle gradient vectors are combined so as to determine a dominant gradient orientation for the target region.

In order to restrain both a deterioration of identification of objects captured by respective capturing apparatuses connected via a network and an increase of a load of the network, a frame rate of the capturing apparatus (proximity camera) is raised according a timing when a person to be captured by the one capturing apparatuses is captured by the other capturing apparatus of performing the capturing next to the one capturing apparatus.

The current document is directed to automated methods and systems, controlled by various constraints and parameters, that identify document sub-images within digital images. Certain of the parameters constrain contour identification and document-subimage-hypothesis generation. The currently described methods and systems identify contours within the digital image, partition the identified contours into four contour sets corresponding to four different regions of the original digital image, construct hypotheses based on these contours for the edges or boundaries of a digital sub-image, and evaluate the hypotheses in order to select a most highly scored hypotheses as a representation of the borders of a digital sub-image within the original received digital image.

A method of label detection includes: obtaining, by an imaging controller, an image depicting a shelf; increasing an intensity of a foreground subset of image pixels exceeding an upper intensity threshold, and decreasing an intensity of a background subset of pixels below a lower intensity threshold; responsive to the increasing and the decreasing, (i) determining gradients for each of the pixels and (ii) selecting a candidate set of the pixels based on the gradients; overlaying a plurality of shelf candidate lines on the image derived from the candidate set of pixels; identifying a pair of the shelf candidate lines satisfying a predetermined sequence of intensity transitions; and generating and storing a shelf edge bounding box corresponding to the pair of shelf candidate lines.

A computer-implemented method for training a normalizing flow. The normalizing flow predicts a first density value based on a first input image. The first density value characterizes a likelihood of the first input image to occur. The first density value is predicted based on an intermediate output of a first convolutional layer of the normalizing flow. The intermediate output is determined based on a plurality of weights of the first convolutional layer. The method for training includes: determining a second input image; determining an output, wherein the output is determined by providing the second input image to the normalizing flow and providing an output of the normalizing flow as output; determining a second density value based on the output tensor and on the plurality of weights; determining a natural gradient of the plurality of weights with respect to the second density value; adapting the weights according to the natural gradient.

A computer-implemented method for explaining an image classifier, the method comprising: receiving an initial image, the initial image having been wrongly classified by the image classifier; receiving an initial gradient of a function executed by the image classifier generated while classifying the initial image, the function being indicative of a probability for the initial image to belong to an initial class; converting the initial image into a latent vector, the latent vector being a representation of the initial image in a latent space; generating a plurality of perturbation vectors using the initial gradient of the function executed by the image classifier; combining the latent vector with each one of the plurality of perturbation vectors, thereby obtaining a plurality of modified vectors; for each one of the plurality of modified vectors, reconstructing a respective image, thereby obtaining a plurality of reconstructed images; transmitting the reconstructed images to the image classifier; for each one of the plurality of reconstructed images, receiving a respective updated gradient of the function executed by the image classifier; using the respective updated gradients, determining amongst the reconstructed images at least one given reconstructed image for which the respective updated gradient is indicative that a new class different from the initial class has been assigned by the image classifier; and outputting the at least one given reconstructed image.

The current document is directed to automated methods and systems, controlled by various constraints and parameters, that identify contours in digital images, including curved contours. Certain of these parameters constrain contour identification to those contours in which the local curvature of a contour does not exceed a threshold local curvature and to those contours orthogonal to intensity gradients of at least threshold magnitudes. The currently described methods and systems identify seed points within a digital image, extend line segments from the seed points as an initial contour coincident with the seed point, and then iteratively extend the initial contour by adding line segments to one or both ends of the contour. The identified contours are selectively combined and filtered in order to identify a set of relevant contours for use in subsequent image-processing tasks.

Techniques are disclosed for extracting micro-features at a pixel-level based on characteristics of one or more images. Importantly, the extraction is unsupervised, i.e., performed independent of any training data that defines particular objects, allowing a behavior-recognition system to forgo a training phase and for object classification to proceed without being constrained by specific object definitions. A micro-feature extractor that does not require training data is adaptive and self-trains while performing the extraction. The extracted micro-features are represented as a micro-feature vector that may be input to a micro-classifier which groups objects into object type clusters based on the micro-feature vectors.

A medical-image processing apparatus according to an embodiment includes processing circuitry. The processing circuit acquires an initial value of an outline corresponding vector that corresponds to an outline of a subject included in medical image data. The processing circuitry updates the outline corresponding vector based on a derivative that is acquired by differentiating a cost function with respect to the outline corresponding vector by the outline corresponding vector, and on the initial value of the outline corresponding vector.