An information processing device performs processing on document image data including first image data to undergo character recognition processing and second image data not to undergo character recognition processing. The information processing device includes a detecting section which detects the first image data, an extracting section which extracts the first image data, and a processing section. The processing section includes a counting section which counts first images, a determining section which determines whether the number of the first images exceeds a threshold, a first performing section which performs first processing when the threshold is exceeded, and a second performing section which performs second processing when the threshold is not exceeded. Through the first processing, the second image is masked with a background color of the document image and character recognition is then performed on the document image. Through the second processing, character recognition is performed on the first images.

In some embodiments, an apparatus includes a memory and a processor operatively coupled to the memory. The processor is configured to receive an electronic document having a set of pages, and partition a page from the set of pages of the electronic document into a set of portions. The processor is configured to convert each portion of the set of portions into a negative image of a set of negative images. The processor is configured to produce, based on an artificial intelligence algorithm, a de-noised negative image of each negative image and convert each de-noised negative image of a set of de-noised negative images into a positive image of a set of positive images, and combine each positive image of the set of positive images to produce a de-noised page. The de-noised page has artifacts less than artifacts of the page of the electronic document.

Embodiments of the disclosure provide systems and methods for generating a report based on medical images of a patient. An exemplary system includes a communication interface configured to receive the medical images acquired by an image acquisition device. The system may further include at least one processor. The at least one processor is configured to receive a user selection of at least one medical image in at least one view. The at least one processor is further configured to automatically generate keywords describing the selected medical image based on a learning network including a convolutional neural network and a recursive neural network connected in series. The at least one processor is also configured to receive a keyword selection among the generated keywords and generate the report based on the keyword selection. The exemplary system additionally includes a display configured to display the selected medical image and the report.

A method disclosed herein uses a processor of a server to function as a processing unit to enhance accuracy of character recognition in a terminal connected to the server, using a communication apparatus of the server. The processing unit may be configured to acquire first data indicating a result of character recognition with respect to image data taken by the terminal. The processing unit can determine a character type of a character included in the image data when it is determined that misrecognition is included in the result of character recognition based on the first data. The processing unit controls the communication apparatus to transmit second data according to the character type to terminal and instructs the terminal to perform character recognition using the second data with respect to the image data in order to improve the accuracy of character recognition.

An image recognition device, an image recognition method and an image recognition unit are capable of performing touch recognition high in accuracy. The image recognition device includes a measurement point determination section adapted to determine a fingertip from an image obtained by a camera, a pattern display section adapted to make a projector display a first pattern having a first linear pattern varying in luminance with a first pitch along a direction parallel to an epipolar line passing through the fingertip, and a second linear pattern varying in luminance with a second pitch along a direction parallel to the epipolar line, and a position detection section adapted to perform touch recognition based on a variation of the first pattern from the image including the first pattern.

An expanding appliance includes a connect port, an image capturing module, an intelligent control module, an image transmitting module, and a result displaying module. The expanding appliance connects an image input device through the connect port, connects a display device through the result displaying module, and connects one or more image-applied function module through the image transmitting module. The intelligent control module generates a demanding command according to a successfully-connected image-applied function module. The image capturing module controls the image input device to capture image data based on the demanding command, and quantizes samples of the image data as computation data. The image transmitting module provides the computation data to the image-applied function module for image identification and receives an identification result. Finally, the intelligent control module triggers the result display module for displaying the identification result on the display device.

A radiation image capturing system includes a radiation image capturing apparatus, an irradiation apparatus and an image processing apparatus. The image processing apparatus generates a first radiation image of a subject based on a signal value generated by the radiation image capturing apparatus with no grid attached irradiated by the irradiation apparatus; performs a low-pass filter process on a pixel value of the first radiation image using a scattering kernel, thereby generating a low frequency image; estimates a body thickness of the subject based on the signal value; estimates a scattered ray content rate based on the body thickness; calculates a scattered ray component in the first radiation image based on the low frequency image and the scattered ray content rate; and subtracts the scattered ray component from the first radiation image, thereby generating a second radiation image.

Methods, systems, and non-transitory computer readable medium are described for automated image measurement for process development and optimization. A method includes receiving an image of a product associated with a manufacturing process; determining, using a trained machine learning model, an image classification for the image; selecting, based on the image classification, one or more image processing algorithms for the image; pre-processing the image based on at least one of the one or more image processing algorithms to generate an enhanced image; measuring, using a first image processing algorithm of the one or more image processing algorithms, one or more attributes of the enhanced image to determine image measurements; and reporting the image measurements. The manufacturing parameters of the manufacturing process are to be updated based on the image measurements.

An eyeball tracking method and apparatus, and a device. The method comprises: acquiring a facial grey-scale image set to be detected (101); judging whether the contour of an eyeball iris is determined in an N-th frame facial grey-scale image in the facial grey-scale image set to be detected (102); if not, detecting an eyeball pupil in the N-th frame facial grey-scale image, and determining the central position of the eyeball pupil in the N-th frame facial grey-scale image (103); in the N-th frame facial grey-scale image, taking the central position of the eyeball pupil as a centre to determine a grey-scale image region corresponding to an eyeball window (104); and according to the grey-scale image region corresponding to the eyeball window, determining the contour of the eyeball iris in the N-th frame facial grey-scale image (105). By judging that the contour of an eyeball iris is not determined in a facial grey-scale image, the tracking of the contour of the eyeball iris can be automatically adjusted, and an eyeball pupil is positioned again. By means of the present invention, the accuracy rate of eyeball tracking is improved, and an eyeball can be automatically identified to detect the central position of an eyeball pupil.

Techniques related to implementing convolutional neural networks for object recognition are discussed. Such techniques may include generating a set of binary neural features via convolutional neural network layers based on input image data and applying a strong classifier to the set of binary neural features to generate an object label for the input image data.