Machine vision-based leaf shape segmentation and statistical method in threshing and redrying air separation process

• Main Authors: SU CHAN, SU YI, YI AOER, XIONG WEN, WANG PENG, NI XUDONG, WU YIQIN, WANG HAI, WU KAI, YUE HENG, ZHU BAOKUN, QIAO DANNA, XU JIAN, ZHAO HUI, YANG YUTING, YU CHUNXIA, CHAI JINFENG, QIAN KUNCHI, JIA YU, LU SHEMING
• Format: Patent
• Language: Chinese; English
• Published: 06.09.2022
• Subjects: CALCULATING; CIGARETTES; CIGARS; COMPUTING; COUNTING; HUMAN NECESSITIES; IMAGE DATA PROCESSING OR GENERATION, IN GENERAL; MANUFACTURE OR PREPARATION OF TOBACCO FOR SMOKING ORCHEWING; PHYSICS; SMOKERS' REQUISITES; SNUFF; TOBACCO

The invention relates to a machine vision-based leaf shape segmentation and statistical method in threshing, redrying and air separation processes. The method comprises the following steps: acquiring a bladeless background image and a bladed real-time image in a wind separation state in a wind separator through a machine vision imaging system, respectively carrying out de-noising processing and enhancement processing on the background image and the real-time image, then carrying out subtraction, carrying out graying on a difference image, then segmenting a blade image by using a threshold method of global fixed blade gray threshold segmentation, and finally obtaining a blade image; the method comprises the following steps: acquiring a binary image of a blade, according to a preset pixel area division standard of an ultra-large piece, a large piece, a middle piece, a small piece and a fragment, counting pixel areas of the ultra-large piece, the large piece, the middle piece, the small piece and the fragment by