The invention discloses a chromosome identification method based on depth learning, belonging to the technical field of chromosome identification. At present, the method of chromosome analysis is basically manual operation, and the examiner needs a lot of training time to master the knowledge of identifying each chromosome type, and the workload is heavier. Even experienced physicians analyze andidentify patients' chromosomes, the process typically takes more than two weeks and takes a longer period of time. And artificial recognition, subjectivity is very strong, easy to be affected by the external environment, the accuracy is not high. As that depth learn method is adopted, the chromosome type can be accurately carry out, Compared with the existing identification technology, the high-efficient identification technology can effectively improve the efficiency of chromosome karyotype analysis, shorten the identification sorting time, complete the automatic classification and sorting ofchromosomes with high accuracy, at the same time, can effectively reduce the work burden of doctors, free from external interference, and the process is simple, reasonable, and can be applied to large-scale outward promotion.