A method performed by a device for occlusion handling in augmented reality is provided. The device can generate at least one pixel classification image in a frame including an occluding object and having foreground, background, and unknown pixels. Generation of the at least one pixel classification image can include (1) calculating an initial foreground pixel probability image, and an initial background pixel probability image, and (2) calculating a normalized depth image based on depth information of the occluding object. The device can obtain an alpha mask to blend a virtual object and the foreground of the at least one pixel classification image based on determining a color of the unknown pixels. The device can render a final composition of an augmented reality image containing the virtual object occluded by the occluding object based on applying the alpha mask to pixels in the at least one pixel classification image.

A method for segmenting images is provided including tessellating an image obtained from one of an image database and an imaging system into a plurality of sectors; classifying each of the plurality of sectors by applying one or more pre-defined labels to each of the plurality of sectors, wherein the pre-defined labels indicate at least one of an image quality metric (IQM) and a metric of structure; assigning each of the plurality of classified sectors an Image Quality Classification (IQC); identifying anchor sectors among the plurality of classified sectors, applying filtering and edge detection to identify target boundaries; applying contouring across contiguous sectors and using the assigned IQC as a guide to complete segmentation of an edge between any two identified anchor sectors; and smoothing across segmented regions to increase parametric second-order continuity.

Computer implemented methods and devices for determining dimensions or distances of head features are provided. The method includes identifying a plurality of features in an image of a head of a person. A real dimension of at least one target feature of the plurality of features or a real distance between at least one target feature of the plurality features and a camera device used for capturing the image is estimated based on probability distributions for real dimensions of at least one feature of the plurality of features and a pixel dimension of the at least one feature of the plurality of features.

The present disclosure relates to image processing. The method includes acquiring at least one of a backward propagation feature of an (x+1)th video frame in a video segment or a forward propagation feature of an (x−1)th video frame in the video segment. The video segment includes N video frames, N being an integer greater than 2, and x being an integer. The method further includes deriving a reconstruction feature of the xth video frame from at least one of the xth video frame, the backward propagation feature of the (x+1)th video frame, or the forward propagation feature of the (x−1)th video frame, and deriving a target video frame corresponding to the xth video frame by reconstructing the xth video frame based on the reconstruction feature of the xth video frame. The target video frame has resolution higher than that of the xth video frame.

A method for performing a computer-aided diagnosis (CAD) includes: acquiring a medical image set; generating a three-dimensional (3D) tumor distance map corresponding to the medical image set, each voxel of the tumor distance map representing a distance from the voxel to a nearest boundary of a primary tumor present in the medical image set; and performing neural-network processing of the medical image set to generate a predicted probability map to predict presence and locations of oncology significant lymph nodes (OSLNs) in the medical image set, wherein voxels in the medical image set are stratified and processed according to the tumor distance map.

The present invention relates to a method of determining stock in an inventory. The method comprises obtaining one or more images comprising one or more objects. Further, estimating a three dimensional (3D) location of a Stock Keeping Unit (SKU) marker associated with each of one or more visible objects. Furthermore, determining a stacking pattern of the one or more objects for each level on the pallet using one of the 3D location of SKU marker and a learning model. Thereafter, detecting at least one of presence or absence of one or more undetected objects at each level based on the stacking pattern and the 3D location of the SKU marker. Finally, determining the stock in the inventory based on the presence or the absence of the one or more undetected objects and the one or more visible objects.

A secondary detection system for integrating automated optical inspection and neural network and a method thereof are disclosed. In the secondary detection system, an automated optical inspection apparatus performs automated optical inspection for pin solder joints on circuit board, and when a detection result indicates abnormal condition, the secondary detection device calculates a detection image probability value based on the component image feature and the template image feature, and calculate pin solder joint image probability values based on the component pin solder joint image feature and the template pin solder joint image feature through siamese neural network, to obtain a minimum probability value among the detection image probability value and pin solder joint image probability values. The minimum probability value is used to determine whether to change the detection result, thereby providing accurate detection result of automated optical inspection and increasing a first pass yield.

This invention relates to a method for training a neural network, comprising detecting a hole in each training image of a plurality of training images; transforming each training image into a transformed image, to suppress non-crack information in the training image; and training a neural network using the transformed images, to detect cracks in images (i.e. in objects in images).

An apparatus for analyzing a payload being transported in a load carrying container of a vehicle is disclosed. The apparatus includes a camera disposed to successively capture images of vehicles traversing a field of view of the camera. The apparatus also includes at least one processor in communication with the camera, the at least one processor being operably configured to select at least one image from the successively captured images in response to a likelihood of a vehicle and load carrying container being within the field of view in the at least one image, and image data associated with the least one image meeting a suitability criterion for further processing. The further processing includes causing the at least one processor to process the selected image to identify a payload region of interest within the image and to generate a payload analysis within the identified payload region of interest based the image data associated with the least one image.

Systems and techniques described herein relate to techniques for improving image detection. In some examples, aspects relate to systems and techniques for improving image detection by performing tracking of objects within captured image frames. A process can include obtaining, from an image capture device, a first image frame including an object. The process can further include determining, using an object detector, an object validation score associated with detection of the object in the first image frame, and determining the object validation score is less than a validation threshold. Based on the object validation score being less than the validation threshold, the process can include tracking the object for one or more image frames received subsequent to the first image frame.