Processing method and host for reducing image detection errors

Abstract

A processing method for reducing image detection errors is provided. The method comprising: classifying a region of interest of an image via an image detection model to find out a pixel scope of a first-type object and another pixel scope of a second-type object; comparing the pixel scope of the first-type object with masks of first-type objects and comparing the pixel scope of the second-type object with masks of second-type objects to get a first determination result and a second determination result, respectively; and determining whether the region of interest contains the second-type object based on the first and the second determination results.

Images