High-speed automatic sheet scanning AI auxiliary interpretation system and sheet scanning method thereof

Abstract

The invention relates to a high-speed automatic film-scanning AI auxiliary interpretation system and a film-scanning method thereof. The high-speed automatic film-scanning AI auxiliary interpretation system comprises a control host and an image acquisition machine. The control host sets scanning parameters of the image acquisition machine table, the scanning parameters comprise the number of a plurality of blocks divided into a slide, the number of a plurality of sub-blocks in the blocks, the position of a central block and a scanning path, and the scanning path is spiral scanning in the clockwise direction or the anticlockwise direction; after the control host selects the scanning path, aligning a focusing position of the image acquisition machine table with the central block on the slide for focusing, and acquiring images of sub-blocks in the central block respectively; then, according to the scanning path, the step of moving the focusing position to the adjacent block for focusing and capturing the image of the sub-block is repeated until all the blocks are scanned. According to the method and the scanning system provided by the invention, the film can be quickly scanned, each image does not need to be refocused, and the focusing times and the consumed time can be greatly reduced.

Description

technical field [0001] The present invention relates to an image processing technology, in particular to a high-speed automatic scanning AI-assisted interpretation system and a scanning method thereof. Background technique [0002] Cell detection is to place the cells in the detection solution, make the test liquid containing the cells into a specimen, and then place the specimen under a microscope. Use high magnification to magnify the specimen and capture the image, and then display the captured image on the computer for inspection. [0003] Each specimen will take 900-1600 images, and the images will not be repeated. In other words, the specimen should be divided into 900-1600 small blocks, and an image should be captured once for each block. If only one set of focal lengths is used for scanning, most of the images may be invalid. If you refocus every time you capture an image, you need to focus 900 to 1600 times, which is too many times. If you refocus every time a f...

AI Technical Summary

Problems solved by technology

If you refocus every time a fixed number of images are captured, although invalid images are reduced, when the lens moves to the edge area and then moves back to the middle area, the back and forth difference is too large and it is easy to fail to pull back the focus. of potentially invalid images

Images