US Patent Application 18193987. IMAGE CAPTURING APPARATUS CAPABLE OF RECOGNIZING RECOGNITION TARGET WITH HIGH ACCURACY, METHOD OF CONTROLLING SAME, AND STORAGE MEDIUM simplified abstract

From WikiPatents
Jump to navigation Jump to search

IMAGE CAPTURING APPARATUS CAPABLE OF RECOGNIZING RECOGNITION TARGET WITH HIGH ACCURACY, METHOD OF CONTROLLING SAME, AND STORAGE MEDIUM

Inventors

Takashi Kobayashi of Tokyo (JP)


IMAGE CAPTURING APPARATUS CAPABLE OF RECOGNIZING RECOGNITION TARGET WITH HIGH ACCURACY, METHOD OF CONTROLLING SAME, AND STORAGE MEDIUM - A simplified explanation of the abstract

  • This abstract for appeared for patent application number 18193987 Titled 'IMAGE CAPTURING APPARATUS CAPABLE OF RECOGNIZING RECOGNITION TARGET WITH HIGH ACCURACY, METHOD OF CONTROLLING SAME, AND STORAGE MEDIUM'

Simplified Explanation

The abstract describes an image capturing device that has an image sensor and a signal processor. The device can switch between a recognition mode and a learning mode. The image sensor captures images, and a sensor recognition section performs recognition processing based on the images. In recognition mode, the signal processor uses a learning model to process the image input from the image sensor section. In learning mode, the signal processor performs machine learning of the learning model based on the recognition result from the sensor recognition section and the image input from the image sensor section.


Original Abstract Submitted

An image capturing apparatus includes an image sensor section and a signal processor for processing an image input from the image sensor section via a signal line. A control arithmetic unit switches the signal processor to a recognition mode or a learning mode. The image sensor section includes an image capturing section for generating the image, and a sensor recognition section for performing recognition processing based on the image. The signal processor includes a recognition section for performing, in the recognition mode, recognition processing by inputting the image input from the image sensor section to a learning model, and a learning section for performing, in the learning mode, machine learning of the learning model based on a recognition result obtained by the sensor recognition section and the image input from the image sensor section.