The present invention proposes a fabric defect detection method based on multi-modal deep learning. First, a tactile sensor is placed onto the fabric surface with different defects to collect the fabric texture images, a camera is used to collect the corresponding fabric external images, and a fabric external image and a fabric texture image constitute a set of fabric detection data; then, a feature extraction network and a multi-modal fusion network are connected to establish a classification model based on multi-modal deep learning, which uses the fabric texture image and fabric external image in each set of collected fabric detection data as input, and the fabric defect as output; said classification model is trained using the collected fabric detection data; finally, the trained classification model is used to detect the fabric defect. The present invention employs vision-touch complementary information, which can greatly improve the accuracy and robustness of detection.

Examples of the disclosure describe systems and methods for identifying, quantifying, and/or characterizing plant stomata. In an example method, a first set of two or more images of a plant leaf representing two or more focal distances is captured via an optical sensor. A reference focal distance is determined based on the first set of images. A second set of two or more images of the plant leaf is captured via the optical sensor, including at least one image captured at a focal distance less than the reference focal distance, and at least one image captured at a focal distance greater than the reference focal distance. A composite image is generated based on the second set of images. The composite image is provided to a trainable feature detector in order to determine a number, density, and/or distribution of stomata in the composite image.

An image marking method, apparatus and system, which relates to the technical field of image processing. The present disclosure includes, when the working mode of the first client is a first mode, receiving a first marking task assigned by a second client, on the condition that the image marking approach is the first marking approach, according to a neural network model, determining a first marking result corresponding to the first original image; on the condition that the image marking approach is the second marking approach, according to an unsupervised algorithm model, determining a second marking result corresponding to the first original image; on the condition that the image marking approach is the third marking approach, receiving a third marking result inputted by a user into the first original image; and sending a target marking result to the second client.

Devices, systems, and methods for real-time food production are disclosed. Extrusion can include including evaluating and controlling one or more production devices to produce desirable food products. Evaluation can be performed by an evaluation system including a convolutional neural network to determine a bulk density value. Control can be performed by a machine learning model on the basis of the bulk density value. Control can include determination of real-time settings for production device parameters.

An image recognition device and method retrieve information on a marker. The marker encodes information identifying an asset by encoding the information in a binary pattern as recesses and non-recesses in the marker. The image recognition is performed by imaging the marker, mapping contrast in the image, identifying variations in the contrast, creating a mesh overlaying the image, identifying the present or absence of recesses from the mesh, and reading the binary pattern represented by the recesses in the marker.

The present invention provides a data collection system comprising: a camera; a location module; a plurality of sensors; and a first processor communicatively coupled to the camera and the location module, the first processor programmed to: obtain a plurality of frames from the camera; obtain a plurality of locations from the location module; obtain a plurality of data measurements from the plurality of sensors; apply a previously trained first neural network model for identifying problematic road segments to frames captured by the camera; and if the first neural network model indicates that a frame is a problematic road segment, save the frame in association with a location provided by the location module.

A hardness tester includes an image acquirer (controller) acquiring an image of a surface (surface image) of a sample captured by an image capturer, an identifier (controller) identifying, based on the surface image of the sample, a non-conformity region inside the image that is unsuitable for the hardness test using predetermined conditions, and a test position definer (controller) defining a test position in an area outside the non-conformity region identified by the identifier.

Provided are various mechanisms and processes for automatic computer vision-based defect detection using a neural network. A system is configured for receiving historical datasets that include training images corresponding to one or more known defects. Each training image is converted into a corresponding matrix representation for training the neural network to adjust weighted parameters based on the known defects. Once sufficiently trained, a test image of an object that is not part of the historical dataset is obtained. Portions of the test image are extracted as input patches for input into the neural network as respective matrix representations. A probability score indicating the likelihood that the input patch includes a defect is automatically generated for each input patch using the weighted parameters. An overall defect score for the test image is then generated based on the probability scores to indicate the condition of the object.

An imaging system can receive an image of a portion of an environment. The environment can include an object, such as a hand or a display. The imaging device can identify a data stream from an external device, for instance by detecting the data stream in the image or by receiving the data stream wirelessly from the external device. The imaging device can detect a condition based on the image and/or the data stream, for instance by detecting that the object is missing from the image, by detecting that a low resource at the imaging device, and/or by detecting visual media content displayed by a display in the image. Upon detecting the condition, imaging device automatically determines a location of the object (or a portion thereof) using the data stream and/or the image. The imaging device generates and/or outputs content that is based on the location of the object.

A system for image analysis for physical defect detection of a storage medium of a first card that comprises an apparatus configured to receive the storage medium of the first card and a first server configured to verify issuance of a second card. The apparatus comprises an optical sensor comprising a light transmitter, a receiver, and a first processor. The optical sensor is configured to produce a light to be directed to the storage medium and to determine that the storage medium comprises a physical defect with the first processor based at least in part upon processed electronic signals. The apparatus further comprises a second processor configured to display outputs requesting authorization to issue the second card. The first server is configured to verify issuance of the second card, wherein the first server is communicatively coupled to the apparatus.