Method for generating reconstructed image

Abstract

The invention provides a method for generating a reconstructed image, which is suitable for an input image with a target object, and comprises the following steps: an encoder converts the input image into a feature map with a plurality of feature vectors; executing a training program according to a plurality of training images of a plurality of reference objects to generate a plurality of feature prototypes associated with the training images and storing the feature prototypes in a memory; selecting a part of the feature prototypes from a memory according to a plurality of similarities between the feature prototypes and the feature vectors; generating an approximate feature map according to the feature prototypes of the part and a plurality of weights, wherein the weights respectively represent the similarity between the feature prototypes and the feature vectors; and the decoder converts the approximate feature map into a reconstructed image. Wherein the encoder, the decoder and the memory form an auto-encoder. The defect classifier realized by applying the image reconstruction method provided by the invention can resist noise in a training data set.

Description

technical field [0001] The present invention relates to defect detection of image-based products, and more particularly, to a method for generating reconstructed images applied to the front-end of defect detection. Background technique [0002] For manufacturers, product appearance assessment is an essential step in quality assurance. Undetected defects such as scratches, bumps and discoloration will increase the cost of returning the product to the factory for repair and loss of customer confidence. Most visual inspection jobs today are still performed by humans because of the difficulty in describing various defects using traditional computer vision algorithms in Automatic Optical Inspection (AOI) machines. However, managing human inspectors has its administrative challenges as it is difficult to maintain consistent inspection standards across different product lines. [0003] Object detector networks have been proposed in the past to solve the above problems. However, ...

AI Technical Summary

Problems solved by technology

Furthermore, many input images are not perfect, and if these imperfect images are excluded, the available standard image samples will be greatly reduced

Images