Abstract:
The invention relates to a method for processing multi-channel image recordings in order to detect hidden objects in a scene, wherein image data from at least two spectrally different camera systems having overlapping fields of vision are subjected to an image data combination. The aim of finding a new possibility for processing multi-channel image recordings for the optoelectronic inspection of persons that achieves an improvement in the detection of hidden objects without the so-called full-body-scanner effect of the millimeter wave scan appearing is achieved according to the invention in that images are recorded as two-dimensional image data in image channels having different frequency ranges, a mask image is produced from the image data of at least one image channel, wherein the image data are differentiated into at least two classes of pixels, the image data of each image channel are transformed into a Gaussian and Laplacian pyramid and the image data of the mask image are transformed into a mask image pyramid adequate therefor, and the image data of the Laplacian pyramids of the individual image channels are combined with the mask image pyramid at pyramid levels that correspond to each other, and then the Laplacian pyramids from corresponding image points of the same pyramid levels are merged into the resulting Laplacian pyramid on the basis of a comparison criterion.
Abstract:
A method (300) for performing image processing utilizes mobile platforms to provide functions including processing, retrieving, annotating and/or mining of media information. According to an exemplary embodiment, the method (300) includes extracting first and second features of an image (320); searching a database according to one of the first and second extracted features and producing candidates in the database that match all or part of the one of the first and second extracted features (340); selecting one or more of the candidates according to the other of the first and second extracted features (350); and enabling display of the selected one or more candidates (350).
Abstract:
The invention concerns a method for creating a saliency map of an image. It comprises the steps of an image comprising the step of hierarchical decomposition of the image into frequential sub-bands, According to the invention, the method comprises the steps of - movement estimation between the current image (I(t)) and a preceding image (I(t-nT)) calculated from the decomposition into frequential sub-bands, estimating a dominant movement for the image and a local movement for each pixel of the current image, - creation of a temporal saliency map obtained from the global and local movement estimation.
Abstract:
Type prediction method, apparatus and electronic device for recognizing an object in an image are disclosed. The method may include processing an image to be processed using a full image recognition technique to obtain a first type prediction result of an object in the image to be processed; processing a subject area of the image to be processed using a feature recognition technique to obtain a second type prediction result of an object of the subject area; determining whether the first type prediction result matches the second type prediction result; if the first type prediction result matches the second type prediction result, determining a type of the object of the image to be processed to be the first type prediction result or the second type prediction result.