Abstract: An abnormality detection device includes an acquisition process unit that acquires plural captured images captured by an imaging device to image an object in different positions or postures in a mode in which a relative positional or postural relationship between the imaging device and the object is specifiable in each of the plural captured images, an extraction process unit that extracts plural abnormality-candidate images in which an abnormality-candidate part is imaged from the plural captured images, the abnormality-candidate part corresponding to a part that possibly includes an abnormality on a surface of the object, an estimation process unit that estimates a depth of the abnormality-candidate part with respect to the surface of the object, and a detection process unit that detects whether the abnormality-candidate part corresponds to abnormal based on the depth of the abnormality-candidate part in each of the plural abnormality-candidate images.

Abstract: A learning device according to an embodiment includes an acquirer, an extractor, a plurality of processors, and an identifier. The acquirer is configured to acquire a paper sheet image which is a captured image of a paper sheet. The extractor is configured to extract a plurality of primary feature images having different objects to be recognized from the paper sheet image acquired by the acquirer. The plurality of processors are configured to perform respective convolution and pooling processes on the plurality of primary feature images extracted by the extractor to generate a plurality of secondary feature images having different objects to be recognized. The identifier is configured to sequentially update and learn a parameter set for identifying a sheet type of the paper sheet on the basis of a result of a combination process on each of the plurality of secondary feature images generated by the plurality of processors.

Abstract: An image processing device that acquires an image from an object includes an illuminator having a plurality of light sources, a sensor unit, and a controller. The plurality of light sources irradiates light of respectively different colors onto an object. The sensor unit outputs luminance data for one line obtained by photoelectrically converting light is reflected from the object of that was irradiated by the illuminator onto the object. The controller acquires a random number, and acquires luminance data for one line output by the sensor unit, while causing one light source selected from the plurality of light sources based on the acquired random number to emit light.

Abstract: According to one embodiment, a sheet processing apparatus includes: an image acquiring device configured to acquire a sheet image that is an image on a sheet; a setting portion configured to set a reference area in the sheet image acquired by the image acquiring device, wherein the setting portion repeats moving the reference area from a starting position to an end position in a first direction by a first moving width, and, upon the reference area reaching the end position, moving the reference area in a second direction intersecting the first direction by a second moving width; a degree of similarity calculating portion configured to extract a feature from an image in the reference area set by the setting portion, and calculate a degree of similarity between the extracted feature and a typical sheet pattern; a moving width calculating portion configured to calculate the first moving width and/or the second moving width based on the degree of similarity calculated by the degree of similarity calculating porti

Abstract: According to one embodiment, an image reading apparatus includes: a first illumination device for irradiating light at a first angle onto a sheet that is being conveyed; a second illumination device for irradiating light at a second angle onto the sheet that is being conveyed; illumination controllers for switching between a first irradiation state in which light is irradiated from the first illumination unit and the second illumination unit onto the sheet and a second irradiation state in which light is irradiated from only either the first illumination device or the second illumination device to the sheet; an imaging device for acquiring a first comparable image in the first irradiation state and for acquiring a second comparable image in the second irradiation state; and a defect detector for detecting a defect of the sheet based on the difference between the first comparable image and the second comparable image.

Abstract: An image processing device that acquires an image from an object includes an illuminator having a plurality of light sources, a sensor unit, and a controller. The plurality of light sources irradiates light of respectively different colors onto an object. The sensor unit outputs luminance data for one line obtained by photoelectrically converting light is reflected from the object of that was irradiated by the illuminator onto the object. The controller acquires a random number, and acquires luminance data for one line output by the sensor unit, while causing one light source selected from the plurality of light sources based on the acquired random number to emit light.

Abstract: According to one embodiment, a postal indicium detection method includes detecting whether or not one of a first face and a second face of a sheet corresponds to a picture card specific face, detecting a postal indicium candidate from the second face based on a first detection result indicating that the first face corresponds to the picture card specific face, and detecting the postal indicium candidate from the first face based on a second detection result indicating that the second face corresponds to the picture card specific face.

Abstract: A defacement degree determination apparatus according to one embodiment of the invention includes an alignment unit which aligns positions of an input image from one printed material printed using a plurality of printing plates with registered images registered in advance, a plate separation processing unit which separates the input image into a plurality of printed components corresponding to the plurality of printing plates, and extracts defacement features from the respective printed components, and a defacement degree determination unit which determines a defacement degree of the printed material based on the defacement features extracted from the respective printed components.

Abstract: According to one embodiment, a barcode processing apparatus comprises a reading unit configured to read an image of a barcode configured by a plurality of bars laid out at a predetermined interval, each bar being formed by a plurality of ink dots, and a barcode recognition unit configured to filter the read image using a mask which has a width smaller than the predetermined interval and larger than a width of one bar, and corresponds to a length of one bar so as to extract a group of a plurality of ink dots, to couple the plurality of ink dots in the extracted group, to detect one bar based on the coupling result, and to recognize a barcode configured by a plurality of detected bars.

Abstract: In a stain determination apparatus or sheet processing apparatus, an input processing unit inputs image information containing an inherent variation that is not related to a stain, a feature extracting unit extracts a plurality of feature information items from the image information input by the input processing unit, a separating unit separates a set of the plurality of feature information items extracted by the feature extracting unit into a inherent variation component and another residual component, a determining unit extracts a main component of stain variation indicating a stain degree in the image information input by the input processing unit from the residual component separated by the separating unit and determines a stain degree in the image information based on the magnitude of the extracted stain variation main component.

Abstract: According to one embodiment, sheet processing includes feature-amount extraction unit configured to extract feature amount from image obtained, standard pattern storage unit configured to prestore standard pattern group for each different soil level, standard pattern group includes standard pattern for each category, selection unit configured to select one standard pattern group from standard pattern groups, similarity calculation unit configured to calculate similarity, based on standard pattern comprised in one standard pattern group selected, and on feature amount extracted, comparison unit configured to compare similarity calculated with preset threshold, determination unit configured to determine category of sheet when similarity is equal to or greater than threshold, as result of comparison, and control unit configured to control selection unit to select one other standard pattern group which has not been selected, when similarity is smaller than threshold, as result of comparison.

Abstract: An information recognition processing apparatus according to an embodiment of the invention includes a storage unit configured to store statistical information indicating results of identification information recognition processing in which an image of a first object to be read is read to acquire a plurality of first read images of different image brightness levels, and recognition of identification information included in the first read image of each brightness level is tried, an image reading unit configured to read an image of a second object to be read to acquire a plurality of second read images of different image brightness levels, and a recognition unit configured to preferentially execute a recognition process of identification information included in the second read image of a predetermined brightness level of the plurality of second read images of the different image brightness levels, based on the statistical information.

Abstract: The class of a sheet is efficiently estimated and a pattern identification process which is robust to a variation in the medium can be performed by dividing an image pattern of the sheet into a plurality of areas (pixels or sets of pixels), weighting and selecting the areas, attaining the identification results for the respective areas and determining the identification result of the whole portion based on a logical combination of the identification results. Particularly, since the area weighting and selecting process is performed based on a difference between the classes and a variation in the class, the calculation amount can be reduced and the identification performance which is higher than that of a method which uniformly processes the whole portion of the pattern can be attained.

Abstract: According to one embodiment of the present invention, there is provided a pattern recognition method of approximating distribution of a set of vectors and a class boundary in a vector space based on basis functions. The method includes defining directional basis functions between two basis vectors, and performing the approximation using a linear combination of the directional basis functions.

Abstract: According to one embodiment, a postal indicium detection method includes detecting whether or not one of a first face and a second face of a sheet corresponds to a picture card specific face, detecting a postal indicium candidate from the second face based on a first detection result indicating that the first face corresponds to the picture card specific face, and detecting the postal indicium candidate from the first face based on a second detection result indicating that the second face corresponds to the picture card specific face.

Abstract: An image processing apparatus according to one embodiment of the invention includes a first image reading unit configured to read an image from an object to be read by visible light, a second image reading unit configured to read an image from the object to be read by invisible light, a first detection unit configured to detect a background image from a first read image read by the first image reading unit, a registration unit configured to register feature information indicative of a feature of the background image detected by the first detection unit, and a second detection unit configured to detect an image indicative of code information from a second read image read by the second image reading unit on the basis of the feature information registered in the registration unit.

Abstract: According to one embodiment, a method of creating a dictionary for soil detection of a sheet includes dividing a sensible area in first and second adjustment images into a plurality of areas, calculating a first characteristic amount of each divided area of the first adjustment image, calculating a second characteristic amount of each divided area of the second adjustment image, calculating a mean and a variance of the first and second characteristic amounts of each area, setting weight data for each area based on the calculated mean and variance, and storing the weight data together with threshold values for soil detection of a sheet.

Abstract: According to one embodiment, a barcode processing apparatus comprises a reading unit configured to read an image of a barcode configured by a plurality of bars laid out at a predetermined interval, each bar being formed by a plurality of ink dots, and a barcode recognition unit configured to filter the read image using a mask which has a width smaller than the predetermined interval and larger than a width of one bar, and corresponds to a length of one bar so as to extract a group of a plurality of ink dots, to couple the plurality of ink dots in the extracted group, to detect one bar based on the coupling result, and to recognize a barcode configured by a plurality of detected bars.

Abstract: An information recognition processing apparatus according to an embodiment of the invention includes a storage unit configured to store statistical information indicating results of identification information recognition processing in which an image of a first object to be read is read to acquire a plurality of first read images of different image brightness levels, and recognition of identification information included in the first read image of each brightness level is tried, an image reading unit configured to read an image of a second object to be read to acquire a plurality of second read images of different image brightness levels, and a recognition unit configured to preferentially execute a recognition process of identification information included in the second read image of a predetermined brightness level of the plurality of second read images of the different image brightness levels, based on the statistical information.

Abstract: In a stain determination apparatus or sheet processing apparatus, an input processing unit inputs image information containing an inherent variation that is not related to a stain, a feature extracting unit extracts a plurality of feature information items from the image information input by the input processing unit, a separating unit separates a set of the plurality of feature information items extracted by the feature extracting unit into a inherent variation component and another residual component, a determining unit extracts a main component of stain variation indicating a stain degree in the image information input by the input processing unit from the residual component separated by the separating unit and determines a stain degree in the image information based on the magnitude of the extracted stain variation main component.