Analysis of DNA molecular genetics features in infected styphnolobium japonicum

摘要

During the study of DNA image recognition at the beginning of infected styphnolobium japonicum, the features of DNA profile are arbitrary. Both characteristic angle and direction change are disorderedly distributed. This study presents a new method to examine the features of early DNA profile based on gray-level co-occurrence matrix (GLCM) and fuzzy mean clustering. Different types of lesion characteristics are distinguished by learning sets and fuzzy means clustering methods. Multilayer cascade classifier is used to extract features of DNA image, and achieve segmentation feature recognition. Simulation results show that the improved method has the efficient ability to evaluate the DNA image with a high rate of infection feature recognition.