A method for automatic defect inspection in wheel shaped casting products is provided, and the three major phases contain preprocessing samples, offline training and online inspection. Specific steps include: collecting and preprocessing training samples, dividing them into three kinds of spoke, rim and axle samples; offline training with the aforementioned three kinds of samples, then generating online detectors respectively for spokes, rims and axles; uploading the well-trained spoke, rim and axle CNN defect detector to the upper computer, placed it in automatic production inspection line. Inspect for defects online automatically. The defect inspection system outputs signals according to the user's requirements. The present invention has a high level of accuracy and reliability, and a strong robustness to variations in illumination, shooting angle and the position of the work piece. It delivers a high level of automation and has no need for an operator to adjust any significant parameters.

Method and system for inspecting hollow glass products of glass product material, wherein the glass products are manufactured by: a. heating the glass product material; b. forming the heated glass product material into at least one glass product; c. cooling the formed glass product; wherein inspecting the glass products comprises the following steps: d. transporting the glass products formed in step b. successively along a predetermined path along at least one infrared light sensitive sensor, wherein with the at least one sensor, of a plurality of the glass products that are transported successively along the at least one sensor, with the at least one sensor per glass product an image is made, wherein step d. is carried out between steps b. and c.; e. processing the images made in step d. for obtaining information about a wall thickness of the glass products.

A method for quantitatively measuring internal defects in an as-cast steel product includes optically scanning at least a portion of a surface of the steel product with a scanning device to create a digital image thereof, analyzing the digital image to calculate a quantitative value for an amount of internal defects therein, and normalizing the quantitative value to a rating according to a standardized scale.

Systems and methods are provided for thermal inspection of tape layup. One embodiment is a method for performing inspection of a tape layup. The method comprises laying up tape onto a surface of a laminate, applying heat to tack the tape to the surface, and generating thermographic images of the tape as applied to the surface.

A three dimensional (3D) detection device has a detection supporter base to be disposed on a transmission device, ultrasonic transceiver modules disposed on at least one inner base surface of the detection supporter base and a controller. When a tested object is transmitted by the transmission device and then enters the detection supporter base, the ultrasonic transceiver modules emit ultrasonic signals to the tested object, and the tested object reflects the ultrasonic signals to the ultrasonic transceiver modules. The ultrasonic transceiver modules generate detection signals according to the reflected ultrasonic signals. The detection signals are sent to the controller, and the controller generates an ultrasonic image corresponding to a tested object according to the detection signals, and then compares the ultrasonic image to a pre-established original 3D image, so to achieve a surface detection objective.

A method of sentencing, accepting or rejecting, a cast component is disclosed. Initially, scanning the component to determine a number of datum points; this can be done using optical scanning techniques. The datum points from the scanned results are then aligned with an ideal design computer aided design (CAD) model of the component. A comparison of the scanned datum points of the component is performed against the data from the ideal design CAD model of the component, and any geometric deviations between the scan and the ideal design CAD model are determined. Using the datum points from the scan of the component an assessment is performed of at least one performance prediction factor for the component. Finally, using dimensional data extracted from the scan and/or the performance prediction factor the component is sentenced for either acceptance or rejection. Additionally, if the component is determined to have a deviation that lies within a pre-determined limit for the dimensional data and/or the performance factor a determination may be made as to whether the component can be reoriented.

The present specification provides a film defect detection system. The film defect detection system may comprise an image acquisition unit configured to acquire an image of a film in a manufacturing process of the film; a defect detection unit configured to detect defects in the film by analyzing the acquired image of the film, using a machine learning algorithm learned to detect a defect in advance, when receiving the acquired image of the film; and an information output unit configured to output information on the defects in the film detected by the defect detection unit.

Method and system for inspecting hollow glass products of glass product material, wherein the glass products are manufactured by: a. heating the glass product material; b. forming the heated glass product material into a glass product; c. cooling the formed glass product; wherein inspecting the glass products comprises the following steps: d. making a plurality of images of the glass product under a plurality of mutually different viewing directions relative to the product with the aid of a plurality of infrared light sensitive sensors, wherein step d. is carried out between steps b. and c.; e. processing in combination of the plurality of images for obtaining at least one parameter that depends on a wall thickness of the glass product.

A method for the non-destructive testing of the volume of a test object, during the course of which a volume raw image of the test object is recorded by a suitable non-destructive imaging testing method. Then, those regions of the volume raw image are identified that are not to be attributed to the test object material. It is checked whether an identified region is completely embedded in regions that are to be associated with the test object material. If necessary, such a region is assimilated to those regions that are to be associated with the test object material, forming a filled volume raw image. Finally, a difference is generated between the volume raw image and the filled volume raw image, forming a first flaw image.

One variation of a method for monitoring manufacture of assembly units includes: receiving selection of a target location hypothesized by a user to contain an origin of a defect in assembly units of an assembly type; accessing a feature map linking non-visual manufacturing features to physical locations within the assembly type; for each assembly unit, accessing an inspection image of the assembly unit recorded by an optical inspection station during production of the assembly unit, projecting the target location onto the inspection image, detecting visual features proximal the target location within the inspection image, and aggregating non-visual manufacturing features associated with locations proximal the target location and representing manufacturing inputs into the assembly unit based on the feature map; and calculating correlations between visual and non-visual manufacturing features associated with locations proximal the target location and the defect for the set of assembly units.