SURF operand-based microscope image splicing method

Abstract

The invention discloses a speeded up robust features (SURF) operand-based microscope image splicing method. In the method, an SURF operand is introduced in the field of microscope image splicing aiming at a large number of spot-like structures such as bubbles, cells and the like in a medical microscope image; the SURF operand is improved; an area is segmented and simultaneously a Gaussian weighting process is simulated; and the number of dimensions of a description vector is halved; therefore, the computational overhead for matching feature points is reduced. By applying the method, the matching accuracy is ensured and simultaneously the speed is improved greatly. The method can be suitable for the field of medical image processing and other computer vision research fields related to image splicing.

Description

technical field

[0001] The invention relates to a method for mosaicing microscope images based on SURF operands. Background technique

[0002] A microscope is a widely used optical instrument. It enlarges the observed object through the function of the optical system, displays the microscopic structure, and understands and studies the characteristics of the object from the microscopic form. At present, it has been widely used in various fields such as biology, pathology, cell histology, and medicinal chemistry. The field of view of the microscope is small, and images of a large field of view cannot be collected at one time. At the same time, for data measurement of large targets, incomplete images need to be spliced ​​into a whole. Therefore, splicing adjacent images together becomes a key link in microscope measurement.

[0003] The image of the microscope shrinks with the increase of the magnification. Under the low magnification lens, the field of view is larger but it ...

Images