Apparatuses, systems, and techniques to facilitate feature detection of a manufactured object such as a PCB using combined images of said manufactured object. In at least one embodiment, an automated optical inspection system (AOI) comprising one or more neural networks can infer based, at least in part, on combined images of a PCB the existence of defects on said PCB.

The embodiments of the Remote Video Inspection System comprise a remote video inspection hardware assembly and a video inspection software system. The remote video inspection hardware assembly is comprised of a camera and a stage assembly. The stage assembly is comprised of a horizontal gimbal, an outer gimbal, and an inner gimbal, a base plate, and a component table. The video inspection software system is comprised of both a local and remote component. The local component is operated by a technician who oversees the remote inspection, and allows for setup, manual positioning, and manual camera adjustment. The remote component is operated by an inspector, and provides video feed to them, as well as a system to control the viewing angle and position of the component/camera.

A training device and a training method for a neural network model. The training method includes: obtaining a data set; completing, according to the data set, a plurality of artificial intelligence (AI) model trainings to generate a plurality of models corresponding to the plurality of AI model trainings respectively; selecting, according to a first constraint, a first model set from the plurality of models; and selecting, according to a second constraint, the neural network model from the first model set.

A target image of a target circuit board and a gold image of a gold circuit board are taken by an image acquisition system. Fiducial points are located on the target image and on the gold image. Perspective transformation is performed on the target image using the fiducial points on the target image for reference and on the gold image using the fiducial points on the gold image for reference. After perspective transformation, an anomalous section of the target image is identified by identifying pixels that have different intensities between the target image and the gold image, the anomalous section being indicative of an unauthorized modification to the target circuit board.

Systems and methods for improved fluid dispensing process control using a machine learning tool are disclosed. In an example method, successive portions of viscous fluid are dispensed by a dispensing device according to operating parameters to train a machine learning tool to associate defect classifications with images of dispensed portions and/or operating parameters associated with dispensing the dispensed portions. The trained machine learning tool is then used in a closed loop fashion in production to detect and correct for defects associated with the dispensed portions to improve quality and production efficiency.

A method for detecting an undesired connection forming a short circuit on a printed circuit board (PCB) based on an original image of the PCB includes steps of: performing binarization on the original image to generate a binary image that has an external boundary; determining, on the binary image, a trace area that corresponds to a trace of the PCB; determining whether the trace area contacts the external boundary of the binary image at four places; and determining that the trace that corresponds to the trace area has an undesired connection forming a short circuit when it is determined that the trace area contacts the external boundary of the binary image at four places.

A substrate work machine for repeatedly performing substrate work, the substrate work machine including a data storing section configured to store component data used in the substrate work, the component data including shape data related to a shape of an electronic component to be mounted on a substrate, a reference value and a tolerance; a data determining section configured to determine whether a difference between measurement data acquired by measuring the electronic component during the substrate work and the reference value of the component data is within a range of the tolerance; a quality information acquiring section configured to acquire work quality information related to a performance condition of a second substrate work machine; and a data correcting section configured to correct at least one of the reference value and the tolerance in accordance with the work quality information.

In one embodiment, a computing system may access design data of a printed circuit board to be produced by a manufacturing process. The system may determine one or more corrections for the design data of the printed circuit board based on one or more correction rules for correcting one or more parameters associated with the printed circuit board. The system may automatically adjust one or more of the parameters associated with the design data of the printed circuit board based on the one or more corrections. The adjusted parameters may be associated with an impedance of the printed circuit board. The one or more corrections may cause the impendence of the printed circuit board to be independent from layer thickness variations of the printed circuit board to be produced by the manufacturing process.

The image data generation device includes a component information input section and an image data generation section. The component information input section is configured to input a mounting position and an outer shape of the electronic component for each of the multiple electronic components to be mounted in the mounting processing. The mounting position includes a position in a height direction orthogonal to a surface of the board. The image data generation section is configured to generate image data for displaying a state when each of the multiple electronic components to be mounted in the mounting processing is arranged at the mounting position of the electronic component based on the mounting position and the outer shape of the electronic component input by the component information input section.

In one embodiment, an X-ray inspection system may nondestructively inspect a printed circuit board to measure a number of dimensions at a number of pre-determined locations of the printed circuit board. The X-ray inspection system may generate a data set for the printed circuit board based on the measured dimensions. The X-ray inspection system may calculate one or more drilling values based on the data set of the printed circuit board. The X-ray inspection system may provide, to a drilling machine, instructions for drilling a number of plated-through vias based on the calculated drilling values for the printed circuit board.