Provided are a damage diagram creation method, a damage diagram creation device, a damage diagram creation system, and a recording medium capable of detecting damage with high accuracy based on a plurality of images acquired by subjecting a subject to split imaging.

In a damage diagram creation method, damage of a subject is detected from each image (each image in a state of being not composed) constituting a plurality of images (a plurality of images acquired by subjecting the subject to split imaging), and thus, damage detection performance is not deteriorated due to deterioration of image quality in an overlapping area. Therefore, it is possible to detect damage with high accuracy based on a plurality of images acquired by subjecting the subject to split imaging. Detection results for the respective images can be composed using a composition parameter calculated based on correspondence points between the images.

A system and method for assisting an insurance service provider to process an insurance request for a roof associated with a building includes the steps of receiving a request for insuring the roof, followed by launching an application for identifying information related to the roof within a selected period of time and utilizing the application to analyze a series of time-lapse images of the roof obtained from past and real-time satellite images of the geographical area. The series of time-lapse images of the roof provides information related to the roof including roof characteristics and other past and present damages and maintenance related information associated with the roof. Comparing sequential changes in a number of pixels in the series of time-lapse images provides the maintenance and damages related information. The insurance service provider compares the above information with preset roof conditions to determine the insurance eligibility status of the roof.

A disclosed system for providing wood-cultivated ginseng age information includes a background sheet made of paper or plastic, a consumer terminal configured to obtain a wood-cultivated ginseng image of a wood-cultivated ginseng placed on the background sheet by a user, receive the start line of a wood-cultivated ginseng rhizome and the end line of the wood-cultivated ginseng rhizome by the user, calculate a wood-cultivated ginseng rhizome length using the obtained wood-cultivated ginseng image and the obtained start line and end line of the wood-cultivated ginseng rhizome, and transmit the calculated wood-cultivated ginseng rhizome length, and a server configured to previously store wood-cultivated ginseng age information corresponding to a wood-cultivated ginseng rhizome length, search for wood-cultivated ginseng age information corresponding to the wood-cultivated ginseng rhizome length received from the consumer terminal, and transmit the retrieved wood-cultivated ginseng age information to the consumer terminal.

The present disclosure describes methods and systems for virtually calibrating geometric sensors with overlapping fields of view. In some embodiments, a geometric sensor may be virtually calibrated by applying a correction value to profile data obtained by the geometric sensor to generate adjusted profile data. The correction factor may be determined based at least in part on X-Y offsets and/or rotational offsets of prior profile data obtained by the geometric sensor relative to corresponding profile data obtained by a reference geometric sensor, and may be recalculated or updated as new sets of profile data are obtained. The adjusted profile data may be used in place of the original profile data in various data processing operations to functionally offset a positional error of the geometric sensor.

A method and a system for patch-based scene segmentation using neural networks are presented. In an embodiment, a method comprises: using one or more computing devices, receiving a digital image comprising test image; using the one or more computing devices, creating, based on the test image, a plurality of grid patches; using the one or more computing devices, receiving a plurality of classifiers that have been trained to identify one or more materials of a plurality of materials; using the one or more computing devices, for each patch of the plurality of grid patches, labelling each pixel of a patch with a label obtained by applying, to the patch, one or more classifiers from the plurality of classifiers; using the one or more computing devices, generating, based on labels assigned to pixels of the plurality of grid patches, a grid of labels for the test image.

An image processing system receives an image depicting a bundle of boards. The bundle of boards has a front face that is perpendicular to a long axis of boards and the image is captured at an angle relative to the long axis. The image processing system applies a homographic transformation to estimate a frontal view of the front face and identifies a plurality of divisions between rows in the estimate. For each adjacent pair of the plurality of divisions between rows, a plurality of vertical divisions is identified. The image processing system identifies a set of bounding boxes defined by pairs of adjacent divisions between rows and pairs of adjacent vertical divisions. The image processing system may filter and/or merge some bounding boxes to better match the bounding boxes to individual boards. Based on the bounding boxes, the image processing system determines the number of boards in the bundle.

A system that includes a computer processor having a plurality of input data devices, a plurality of output data devices, and a plurality of sensors; and a mechanical assembly integrated with the computer processor to reposition a piece of wood lumber based on software code executing in the computer processor. In some embodiments, the system performs a method that includes eliciting and receiving into the computer processor data parameters from a first human user; obtaining incoming data points about the lumber from the plurality of sensors; processing and storing the data parameters; comparing the incoming data points to the data parameters to obtain comparison results; and, based on the comparison results, (1) rejecting the lumber to a preprogrammed position, (2) feeding the lumber into a saw assembly as positioned, or (3) repositioning the lumber to a more optimal position prior to feeding the lumber to the saw assembly.

A method of board lumber grading is performed in an industrial environment on a machine learning framework configured as an interface to a machine learning-based deep convolutional network that is trained end-to-end, pixels-to-pixels on semantic segmentation. The method uses deep learning techniques that are applied to semantic segmentation to delineate board lumber characteristics, including their sizes and boundaries.

The present disclosure describes methods and systems for virtually calibrating geometric sensors with overlapping fields of view. In some embodiments, a geometric sensor may be virtually calibrated by applying a correction value to profile data obtained by the geometric sensor to generate adjusted profile data. The correction factor may be determined based at least in part on X-Y offsets and/or rotational offsets of prior profile data obtained by the geometric sensor relative to corresponding profile data obtained by a reference geometric sensor, and may be recalculated or updated as new sets of profile data are obtained. The adjusted profile data may be used in place of the original profile data in various data processing operations to functionally offset a positional error of the geometric sensor.

Image feature alignment is provided. In some implementations, a computer-readable tangible medium includes instructions that direct a processor to access a reference feature point associated with a high contrast region in a first sub-image that is associated with a first section of a borehole. Instructions are also present that direct the processor to identify several candidate feature points in a second sub-image associated with a second section of the borehole adjacent to the first section of the borehole, with each of the candidate feature points being believed to possibly be associated with the high contrast region. Additional instructions are present that direct the processor to prune the candidate feature points using global solution pruning to arrive at a matching candidate feature point in the second sub-image.