A semiconductor-inspection tool scans a semiconductor die using a plurality of optical modes. A plurality of defects on the semiconductor die are identified based on results of the scanning. Respective defects of the plurality of defects correspond to respective pixel sets of the semiconductor-inspection tool. The scanning fails to resolve the respective defects. The results include multi-dimensional data based on pixel intensity for the respective pixel sets, wherein each dimension of the multi-dimensional data corresponds to a distinct mode of the plurality of optical modes. A discriminant function is applied to the results to transform the multi-dimensional data for the respective pixel sets into respective scores. Based at least in part on the respective scores, the respective defects are divided into distinct classes.

A cosmetic testing device for electronic devices that uses the electronic device's own camera to take photographs of the electronic device in order to determine its cosmetic condition.

Invention refers to quality control system for detection and identification of insects, pests and the like in such product like cereals, seed, grains and the like. The Quality control system comprises a sample collecting device (1) for collecting a sample in pre-determined amount from a product (P) to be tested, a self-cleaning cyclone filter (2) for cleaning the sample, a vibro separator (3) for separation of the sample in pre-determined sizes or grades, a heating chamber (4) for heating the sample separated in the vibro separator. The system further comprises an identification device (5) for detection and identification of live insects, pests and the like in said sample.

A product defect detection method, device and system are disclosed. The method comprises: acquiring a sample image of a product, extracting candidate image blocks probably including a product defect from the sample image, and extracting preset shape features corresponding to the candidate image blocks and texture features corresponding to the candidate image blocks; training a first-level classifier using the preset shape features to obtain a first-level classifier that can further screen out target image blocks probably including a product defect from the candidate image blocks; training a second-level classifier using the texture features to obtain a second-level classifier that can correctly identify a product defect; and when performing product defect detection, inputting preset shape features of candidate image blocks extracted from a product image into the first-level classifier, and then inputting texture features of obtained target image blocks into the second-level classifier to detect a defect in the product.

The invention discloses a method for identifying frostbite condition of grain seeds using the spectral feature wavebands of the grain seed embryo hyperspectral images. At first, hyperspectral image of the grain seed in the embryo side is collected, and the hyperspectral image in the embryo region of the grain seed is obtained. Then the average spectra is calculated and the wavebands containing noise are eliminated, the spectral feature wavebands are extracted by using related algorithm and the spectra value corresponding with the waveband is obtained. Next, the feature waveband spectral value and the category label of the known frostbite category grain seed are input into the classification model for obtaining the optimal training classification model. Finally, the feature waveband spectral value of each unknown frostbite category grain seed is input into the established model, the frostbite condition of seed is identified.

The present invention provides an automated optical inspection and classification apparatus based on a deep learning system, comprising a camera and a processor. The processor executes a deep learning system after loading data from a storage unit and the processor, and comprises an input layer, a neural network layer group, and a fully connected-layer group. The neural network layer group is for extracting to an input image and thereby obtaining a plurality of image features. The fully connected-layer group is for performing weight-based classification and outputting an inspection result.

There is presented a system and method for detecting mobile device fault conditions, including detecting fault conditions by software operating on the mobile device. In one embodiment, the present invention provides for systems and methods for using a neural network to detect, from an image of the device, that the mobile device has a defect, for instance a cracked or scratched screen. Systems and methods also provide for, reporting the defect status of the device, working or not, so that appropriate action may be taken by a third party.

A method for adjusting a condition for determining a quality of an inspection object comprises: acquiring measurement values for the structures of a plurality of inspection objects; determining whether each of the plurality of inspection objects is good or defective by comparing error values of the measurement values with respect to design values with a predetermined reference value; identifying one or more inspection objects in which determination error has occurred among the plurality of inspection objects; generating and outputting an inspection result graph including the number of inspection objects according to the error values, the reference value, and the number of the one or more inspection objects in which the determination error has occurred; updating the reference value according to a graphical input; and redetermining whether each of the plurality of inspection objects is good or defective by comparing the error values with the updated reference value.

A product defect detection method, device and system are disclosed. The product defect detection method comprises: constructing a defect detection framework including a classification network, a locating detection network and a judgment network; training the classification network by using a sample image of a product containing different defect types to obtain a classification network capable of classifying the defect types existing in the sample image; training the locating detection network by using a sample image of a product containing different defect types to obtain a locating detection network capable of locating a position of each type of defect in the sample image; inputting an acquired product image into the defect detection framework, inputting a classification result and a detection result obtained into the judgment network to judge whether the product has a defect, and detecting a defect type and a defect position when the product has a defect.

A belt examination system includes a defect candidate detecting processor that detects a candidate for a belt defect that is an abnormal portion of an intermediate transfer belt of an image forming apparatus from a belt image that is an image of the intermediate transfer belt, the defect candidate detecting processor executes a background pattern reduction step to reduce a texture-pattern like background noise present in the belt image and detects the candidate based on the belt image generated during the background pattern reduction step, and the background pattern reduction step is to replace, in the belt image, a color value within a specific range of color values not including a lowest color value of the belt defect with a specific color value within the specific range.