Method for identifying non-inspectable objects in packaging, and apparatus and storage medium applying method

An object identification apparatus for identifying target objects in a sealed case;

• the object identification apparatus comprises a processor and a storage;
• the processor executes program codes stored in the storage to implement the following steps:
• a modeling module, configured to model a detection model by using standard data as a training set;
• the standard data comprises sample images of the sealed case with different number and locations of the target objects;
• a capacity of the sealed case is M, and a number of the sample image is less than 2M;
• an image capturing module, configured to capture an image of the sealed case by a camera as a detection image;
• an image processing module, configured to divide the detection image into a specified number of the detection regions in a predetermined manner;
• an extracting module, configured to extract images of first detection regions in the detection image as local images, and input the local images into the detection model;
• and
• a statistics module, configured to obtain number information and location information of the target objects in the sealed case based on a statistic result outputted by the detection model.

A storage medium comprises a storage medium and a processor;

• the storage medium stores instructions being executed by the processor to implement following steps:
• modeling a detection model by using standard data as a training set;
• the standard data comprises sample images of the sealed case with different number and locations of the target objects;
• a capacity of the sealed case is M, and a number of the sample image is less than 2M;
• capturing an image of the sealed case by a camera as a detection image;
• dividing the detection image into a specified number of the detection regions in a predetermined manner;
• extracting images of first detection regions in the detection image as local images;
• inputting the local images into the detection model;
• and
• obtaining number information and location information of the target objects in the sealed case based on a statistic result outputted by the detection model.