Observation of Animal and Plant Microstructure Based on Computer Vision

  • Xie Jingwei
Keywords: Animal and Plant, Chromosome, Microstructure, Cell Count.


Microscopic observation and counting of plant cells play an important role in many fields such as botanical
images. The traditional manual counting method is inefficient, time-consuming and low accuracy. At present,
the research of automatic cell counting by image processing technology is mostly a few cell adhesion, cell shape
integrity and size uniformity. In fact, most of the cell images are irregular, uneven and conglutinated. In this
paper, double threshold detection algorithm is used to segment and count the adherent cells. The accuracy of the
algorithm is high, which can meet the actual experimental requirements. The threshold set in this paper can be
considered to add other sample characteristics, and the appropriate threshold can be obtained by training
different cell samples. In order to improve the accuracy, the classification algorithm was used to classify the
adherent cells. In addition, cell segmentation algorithm can be based on edge detection, threshold and region
algorithm to get more adaptive segmentation algorithm. It has certain reference value for the observation of
plant microstructure.