This disclosure relates to system and method for optical character recognition. In one embodiment, the method comprises providing an image data to a plurality of customized machine learning algorithms or various customized neural networks, configured to recognize a set of pre-defined characters. The method comprises presenting one or more suggestions for the character to the user in response to negative character recognition, and training a customized machine learning algorithm corresponding to the character if one of the suggestions is identified by the user. If the suggestions are rejected by the user, the method comprises prompting the user to identify the character and determining presence of the character in the set of pre-defined characters. The method further comprises training a customized machine learning algorithm corresponding to the character if the character is present, or dynamically creating a customized machine learning algorithm corresponding to the character if the character is not present.

A signal processing apparatus includes an acquisition unit that acquires input data and detection target data, a noise strength setting unit that sets a noise strength K used to a predetermined stochastic resonance processing and a stochastic resonance processing unit that performs the predetermined stochastic resonance processing and outputs processed data. The predetermined stochastic resonance processing is a processing based on a formula in which processed data J(x) is represented by I(x), the noise strength K and the threshold value T and the processed data J(x) corresponds to a result in a case where M is infinite in the following formula,

[00001]$J\left(x\right)=\frac{1}{M}.Math.\underset{m=1}{\overset{M}{.Math.}}.Math.j\left(x,m\right).$

The noise strength setting unit sets the noise strength based on a function of a correlation coefficient between the result of the predetermined stochastic resonance processing and the detection target data and the noise strength K.

A system and method for detecting counterfeit document. The invention evaluates identifying the Region of Interest (ROI) where patterns such as VOID/COPY Pantographs are located; forming an image of the ROI; cropping the image of ROI; applying multichannel filtering for texture/pattern analysis; detecting object/edges; partitioning a group of data points into clusters; converting gray scale image to binary form using thresholding; applying motion blur to the binary image and further applying thresholding; determining bounding area of bounding boxes to make the identified characters machine readable for OCR. A counterfeit document is detected where the characters under pattern (e.g. VOID/COPY Pantograph) are not detected.

An image processing apparatus has a stochastic resonance processing unit executing a stochastic resonance processing to obtain a result. The result corresponds to a result that is calculated in a case where each of a plurality of pixel signals constituting reading image data is added noise and subjected to a binary processing and a plurality of results obtained by parallelly performing above step are synthesized and the parallel number is infinite. The stochastic resonance processing unit sets, with regard to a pixel signal as a processing target among the plurality of pixel signals, at least one of a strength of the noise and a threshold value used for the binary processing based on a pixel signal of the input image data corresponding to the pixel signal.

Apparatus, methods, and articles of manufacture for implementing crowdsourcing pipelines that generate training examples for machine learning expression classifiers. Crowdsourcing providers actively generate images with expressions, according to cues or goals. The cues or goals may be to mimic an expression or appear in a certain way, or to “break” an existing expression recognizer. The images are collected and rated by same or different crowdsourcing providers, and the images that meet a first quality criterion are then vetted by expert(s). The vetted images are then used as positive or negative examples in training machine learning expression classifiers.

An image processing system includes a reading unit, a checking unit, a display control unit, and an image correcting unit. The reading unit reads image information from an original manuscript to generate image data. The checking unit checks whether or not a plurality of items listed in the original manuscript include an unentered item with the result of a character recognition process for recognizing a character contained in the image data. When the unentered item is present, the display control unit performs control to display information for informing a user of the unentered item. The image correcting unit adds a text with respect to the unentered item according to an input made by a user to correct the image data.

A method and non-transitory computer-readable medium capture an image of bulk grain and apply a feature extractor to the image to determine a feature of the bulk grain in the image. For each of a plurality of different sampling locations in the image, based upon the feature of the bulk grain at the sampling location, a determination is made regarding a classification score for the presence of a classification of material at the sampling location. A quality of the bulk grain of the image is determined based upon an aggregation of the classification scores for the presence of the classification of material at the sampling locations.

The present invention relates to an electronic device comprising a capacitive fingerprint sensor having a fingerprint sensing surface forming part of an exterior surface of the electronic device; a protective film arranged on the exterior surface of the electronic device comprising the fingerprint sensing surface; wherein the protective film is arranged to at least partially cover the fingerprint sensing surface; and wherein the protective film comprises an electrically conductive pattern covering the fingerprint sensing surface, the electrically conductive pattern being configured to enable verification of a functionality of the fingerprint sensor. The invention also relates to a method of manufacturing an electronic device comprising a fingerprint sensor, and to a method of verifying a functionality of a fingerprint sensor.

A mobile device with camera automatically captures an image of a check from a video sequence. A computing application assesses quality metrics of a frame of the video and, if acceptable, initiates capture of the check in that frame without user selection. Metrics include an aspect ratio of the check, image quality of the routing transit symbols that delineate a routing transit number on a MICR line of the check, distances between the routing transit symbols and to an edge of the check, recognition of digits of the routing transit number, checksum of the routing transit numbers, and image sharpness. Other embodiments note cropping of the check from the background of the image, properly orienting the check for viewing, and providing color coded visual feedback to users about the quality of the image frame about the check, to name a few.

The present invention relates to a license plate number input device and a method therefor. The license plate number input device according to the present invention comprises: a characteristic information extraction unit for extracting characteristic information on a license plate from a vehicle image; a keypad provision unit for providing at least one input keypad on the basis of the extracted characteristic information; and an out unit for displaying input information input by a user through the provided input keypad.

The license plate number input device according to the present invention is expected to have the effect of allowing a user to input license plate number quickly and easily by providing an image of a license plate region and an input keypad on the basis of characteristic information extracted from a vehicle image.

In particular, the license plate number input device allows the license plate in the vehicle image to be more easily compared with a virtual license plate and thus is expected to have the effect of allowing easy verification while improving naked-eye discrimination of the license plate in the vehicle image.