Video microscopic image cell automatic tracking method based on Meanshift arithmetic

Abstract

The invention discloses a video microscopic image cell automatic tracking method based on Meanshift arithmetic, mainly solving the problem that cells cannot be tracked accurately in special conditions in the prior art. The method comprises the steps of (1) carrying out binary segmentation on an input image and extracting cell information; (2) establishing an uniform motion model for cells to be tracked and initializing; (3) clustering the binary images of the next moment by using the Meanshift arithmetic to obtain an optimum prediction position; (4) calculating cost function values between the cells to be tracked and questionable matching cells; (5) adopting the questionable matching cell with the smallest cost function value as the best matching cell of the cell to be tracked; (6) updating state vector and error covariance matrix in the step (2); and (7) circulating from the step (3) to the step (6) till the last frame of image. The invention has the advantages of accurate cell prediction position confirmation and good tracking effect in special conditions and can be used for medical image processing.

Description

technical field [0001] The invention belongs to the field of digital image processing, and relates to a method for automatically tracking cells in video microscopic images, which can be used for cell movement analysis. Background technique [0002] It is very important to study the movement of cells. From the beginning of the establishment of cytology, the movement of cells was the first to be studied. At that time, due to the constraints of experimental equipment, the microscopic level that the human eye can see is only the whole cell under a microscope, unlike Today, under a high-power microscope, the basic structure of cells can be observed, so at that time cytological research focused more on the movement of cells. Times have changed, and the development of science and technology today makes scientific research convenient and effective. Although more The scientific research activities of the company have penetrated into the more microscopic field of cells, but the import...

AI Technical Summary

Problems solved by technology

[0004] At present, the domestic method of cell tracking is mainly manual timing detection and recording with the aid of microscopic equipment, which has a low degree of automation. The complexity and variability of the form, such as cell division, combination, aggregation, intersection and noise and interference of the video image itself, make the manual observation process more difficult and have a higher error rate

[0005] In recent years, foreign research institutions have achieved some research results in automatic cell tracking, and developed a batch of automatic cell tracking systems with the help of image analysis software. For example, recently, a British research team composed of biologists and computer software development experts developed In a breakthrough study, researchers at Rice University designed a new computer image analysis system that can track cells living in living tissue models. Three-dimensional individual cancer cells, but they are still immature in terms of commercialization. Although these devices can calculate the position of cells on the image through image processing software and analysis software, but for cell division, cell aggregation and multi-cell movement trajectory Difficult to deal with special cases such as crossover

Images