Hon Hai Precision Industry Co., Ltd Patent Portfolio Statistics

Publication number: US20220189008A1
Application number: 17/551,477

Abstract:
A method for detecting data defects and a computing device applying the method obtains a test image for analysis. A field to which the test image relates is determined. Based on the field, a target convolutional layer is determined from a convolutional neural network. The target convolutional layer is used to extract features of the test image. A target score of the test image and a score threshold corresponding to the field are determined. If the target score is less than the score threshold, it is determined that the test image reveals defects, thereby improving an accuracy of defect detection.

Publication date: 2022-06-16
Applicant: Hon Hai Precision Industry Co., Ltd.
Inventors: Chin-Pin Kuo

---

Publication number: US20220093886A1
Application number: 17/541,650

Abstract:
A polymer solar cell includes a photoactive layer, a cathode electrode, and an anode electrode. The photoactive layer includes a polymer layer and a carbon nanotube layer. The polymer layer includes a first polymer surface and a second polymer surface opposite to the first polymer surface. A portion of the carbon nanotube layer is embedded in the polymer layer, and another portion of the carbon nanotube layer is exposed from the polymer layer. The cathode electrode is located a surface of the carbon nanotube layer away from the polymer layer. The anode electrode is located on the first polymer surface and spaced apart from the carbon nanotube layer. The entire second polymer surface is exposed.

Publication date: 2022-03-24
Applicant: Hon Hai Precision Industry Co., Ltd.
Inventors: Liu Peng

---

Publication number: US20220189193A1
Application number: 17/537,987

Abstract:
A method for detecting and counting pedestrians in real-time for statistical purposes together with facial recognition of such pedestrians acquires video for analysis. Images showing pedestrians are extracted, such pedestrians being identified by a first detection model and pedestrian frames are outputted. A facial identification operation is executed based on the pedestrian frames and facial sub-frames are extracted. Removal of duplications of individual pedestrians in the images in the facial sub-frames is executed by a second detection model. Path of movement of detected pedestrians is tracked and labeled based on a specified algorithm when the faces of individual pedestrians are not recorded in a database. When the path of movement passes through a specified location, the face-imaged pedestrian is considered a target object, a total number of the target objects is counted. An apparatus applying the method is also disclosed.

Publication date: 2022-06-16
Applicant: Hon Hai Precision Industry Co., Ltd.
Inventors: Wang Wei-Chun

---