Bacterial colony automatic detection method

Abstract

The invention discloses a bacterial colony automatic detection method, belonging to the fields of biomedicine image processing, environment monitoring and the like. The method comprises the following steps:inputting bacterial colony images; carrying out image gradation processing and pre-processing on the bacterial colony images, and acquiring the processed bacterial colony images; carrying out ideal bacterial colony image reconstruction according to the processed bacterial colony images; and post-processing the ideal bacterial colony images, and detecting the bacterial colonies, wherein, when the bacterial colony images are acquired at different conditions, the input bacterial colony images are converted into the ideal bacterial colony images by eliminating illumination deviation, and the ideal bacterial colony images are subject to post-processing, thus the bacterial colonies are detected. The bacterial colony automatic detection method does not rely on a specific image acquiring device, bacterial colony types and external environmental conditions, and the image segmentation and bacterial colony extracting processes also do not rely on experimental threshold value or prior knowledge, Thus the bacterial colony automatic detection method has the advantages that the universality is high, the application range is enlarged; and large-scale experiments prove the accuracy and precision of the bacterial colony automatic detection method.

Description

technical field [0001] The invention relates to the fields of biomedical image processing, environmental monitoring and the like, in particular to an automatic detection method for bacterial colonies. Background technique [0002] With the development of computer science and technology, computing technology has penetrated into various fields of other disciplines, especially biomedicine, environmental monitoring and other fields. Because computers have strong computing and storage capabilities, they have unique advantages in the acquisition, processing, transmission and storage of digital signals. Therefore, the interdisciplinary research between the fields of biomedical image processing and information science can greatly promote the discovery of biomedical knowledge and the improvement of research methods, and the interdisciplinary research between the field of environmental detection and information science can promote modern detection. Research and development of methods...

AI Technical Summary

Problems solved by technology

However, researchers often need to spend a lot of time to prepare a large number of samples for training models, and the drastic changes in colony color, texture and shape will bring many difficulties to the automatic colony detection method for model training; 2) see the colony detection for image segmentation [5-9] , the existing methods usually choose the global threshold method, the block-based local threshold method or the threshold selection method based on specific domain knowledge

Since the optimal threshold selection is often data-dependent, such methods have great limitations in large-scale applications.

Images