Introducing the features of DWORLD DVS.
technology for detecting a desired object (defect) with respect to an input image.
Image preprocessing process using convolution to remove input noise and unnecessary information to suit the detection purpose.
by learning only the normal image, detect an outlier that deviates from the normal image level and it is judged as a defect.
It can be used when the characteristics of the detected defect are diverse and it is not easy to obtain an image of the defect.
Good for image analysis as a type of artificial neural network
Train a deep learning neural network using normal images and bad images
Learn about defects to determine whether the input image is normal and detect the defect area