CN-121982062-A - Droplet flow rate determination method, terminal device and computer program product
Abstract
The application is applicable to the technical field of computers, and provides a droplet flow rate determining method, terminal equipment and a computer program product, wherein the method comprises the steps of obtaining a continuous image sequence of droplets; the method comprises the steps of inputting a continuous image sequence into a target detection model for processing to obtain drop information of each drop in each frame of drop image in the continuous image sequence, determining a drop combination related to a cross frame based on the drop information of each drop in each frame of drop image, and calculating to obtain speed information of the drop combination based on the drop information of the drop combination. The application can accurately extract the droplet information through continuous image sequence and model processing, and can also identify the droplet information of any droplet in real time to determine the droplet associated with the cross frame, thereby accurately determining the velocity of the droplet, and solving the problems that the detection is inaccurate under the conditions of illumination change, background noise and droplet deformation, the detection is easy to be missed and false in high-density and high-speed scenes, and reliable real-time tracking is difficult to realize in the prior art.
Inventors
- CHEN YUCHAO
- ZHANG FENG
Assignees
- 利德健康科技(广州)有限公司
Dates
- Publication Date
- 20260505
- Application Date
- 20251225
Claims (10)
- 1. A method of determining a droplet flow rate, comprising: Acquiring a continuous image sequence of the liquid drops; Inputting the continuous image sequence to a trained target detection model for processing to obtain the droplet information of each droplet in each frame of droplet image in the continuous image sequence; determining a cross-frame associated droplet combination based on droplet information of each droplet in each frame of droplet images, wherein the droplet combination comprises a first droplet and a second droplet which are the same droplet in any two frames of adjacent images; and calculating the speed information of the liquid drops based on the liquid drop information of the liquid drop combination.
- 2. The drop flow rate determination method of claim 1, wherein determining a drop combination associated across frames based on drop information for each drop in the respective frame of drop images comprises: Acquiring adjacent frame images adjacent to the image where the first liquid drop is located in the continuous image sequence; Determining a plurality of candidate droplets associated with the first droplet in the adjacent frame image in a cross-frame manner based on the droplet information of the first droplet, the droplet information of each initial droplet in the adjacent frame image, and a candidate matching rule; determining the second droplet based on the droplet information of the first droplet, the droplet information of the plurality of candidate droplets, and a cross-frame correlation rule to obtain the droplet combination.
- 3. The method of determining a droplet flow rate according to claim 2, wherein the droplet information includes droplet coordinate information and droplet size information, wherein the determining a plurality of candidate droplets in the adjacent frame image associated with the first droplet across frames based on the droplet information of the first droplet, the droplet information of each initial droplet in the adjacent frame image, and a candidate matching rule includes: Calculating to obtain a coordinate distance between the first droplet and each initial droplet based on the droplet coordinate information of the first droplet and the droplet coordinate information of each initial droplet; Calculating a size change ratio between the first droplet and each initial droplet based on the droplet size information of the first droplet and the droplet size information of each initial droplet; And aiming at any initial liquid drop, if the coordinate distance corresponding to the initial liquid drop is smaller than a set distance and the size change proportion corresponding to the initial liquid drop is smaller than a set proportion, determining the initial liquid drop as the candidate liquid drop.
- 4. The drop flow rate determination method of claim 3, wherein the determining the second drop based on drop information of the first drop, drop information of a plurality of the candidate drops, and a cross-frame correlation rule comprises: determining a target candidate liquid drop with the minimum coordinate distance and the minimum size change proportion from a plurality of candidate liquid drops; determining a flow direction between the first droplet and the target candidate droplet based on the droplet coordinate information of the first droplet and the droplet coordinate information of the target candidate droplet; and if the flow direction accords with a set direction, determining the target candidate liquid drop as the second liquid drop.
- 5. The method of any one of claims 1-4, wherein the droplet flow rate determination information includes droplet coordinate information, the velocity information includes instantaneous velocity, and the calculating the velocity information of the droplets based on the droplet information of the droplet combination includes: Calculating displacement variation of the same liquid drop in any two frames of adjacent images based on the liquid drop coordinate information of the first liquid drop and the liquid drop coordinate information of the second liquid drop; And calculating the instantaneous speed of the same liquid drop in the adjacent images of any two frames based on the displacement variation and the time interval between the adjacent frames corresponding to the continuous image sequence.
- 6. The method for determining a flow rate of droplets according to claim 5, wherein calculating the instantaneous velocity of the same droplet in the adjacent images of any two frames based on the displacement variation and the time interval between the adjacent frames corresponding to the continuous image sequence includes: Calculating an actual physical distance corresponding to the displacement variation based on the displacement variation and a space calibration parameter corresponding to the continuous image sequence; and calculating the instantaneous speed based on the actual physical distance and the time interval.
- 7. The method of determining a flow rate of droplets according to any one of claims 1 to 4, wherein the droplet assembly includes respective sub-assemblies of a plurality of image groups within a set acquisition time period, each of the sub-assemblies including a first droplet and a second droplet which are the same droplet, the first droplet of each of the sub-assemblies being different, each of the image groups including any two adjacent images, the droplet information including droplet coordinate information including an average velocity corresponding to the set acquisition time period, and the calculating the velocity information of the droplet based on the droplet information of the droplet assembly includes: Calculating displacement variation of the same liquid drop corresponding to each sub-combination in the arbitrary two frames of adjacent images based on the liquid drop coordinate information of the first liquid drop and the liquid drop coordinate information of the second liquid drop corresponding to each sub-combination; Calculating to obtain respective instantaneous speeds of the sub-combinations based on the displacement variation corresponding to the sub-combinations and the time intervals between adjacent frames corresponding to the continuous image sequences; and carrying out average value calculation on the instantaneous speeds corresponding to the sub-combinations to obtain the average speed corresponding to the set acquisition time period.
- 8. The method of determining a flow rate of droplets according to claim 1, wherein the droplet combinations include a plurality of sub-combinations each including a first droplet and a second droplet which are in a same image group, each of the sub-combinations including a first droplet and a second droplet which are a same droplet, each of the first droplets of the sub-combinations being different, the same image group including any two adjacent images, at least one of the any two adjacent images of each of the droplet combinations being different, the droplet information including droplet coordinate information including an average velocity corresponding to the continuous image sequence, the calculating the droplet velocity information based on the droplet information of the droplet combinations includes: obtaining a time sequence track of each liquid drop in the continuous image sequence based on the liquid drop coordinate information of each liquid drop combination and the time interval between adjacent frames corresponding to the continuous image sequence, wherein the time sequence track comprises the coordinate information of the liquid drop at different times; calculating the average speed of each liquid drop based on each time sequence track; And carrying out average value calculation on the average speed of each liquid drop to obtain the average speed corresponding to the continuous image sequence.
- 9. A terminal device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that the processor implements the droplet flow rate determination method according to any one of claims 1 to 8 when executing the computer program.
- 10. A computer program product comprising a computer program which, when run, implements the droplet flow rate determination method according to any one of claims 1 to 8.
Description
Droplet flow rate determination method, terminal device and computer program product Technical Field The application belongs to the technical field of computers, and particularly relates to a liquid drop flow rate determining method, terminal equipment and a computer program product. Background In practical application, the microfluidic droplet technology is widely applied to the fields of single cell analysis, antibody screening, molecular diagnosis and the like, and the flow velocity of droplets in a microfluidic channel is an important parameter affecting detection stability, separation efficiency and system performance. Thus, obtaining accurate real-time measurements of droplet flow rates is a central need for related devices. The prior art generally recognizes the droplet position and estimates the velocity through algorithms such as threshold segmentation, edge detection, etc., after capturing an image of the droplet using a high-speed camera. However, the prior art is sensitive to illumination, noise and droplet morphology change, and is easy to miss detection and misdetection in a high-density and high-speed scene, so that reliable real-time tracking is difficult to realize. Disclosure of Invention The embodiment of the application provides a liquid drop flow rate determining method, terminal equipment and a computer program product, which are used for solving the problems that the prior art is sensitive to illumination, noise and liquid drop form change, missed detection and false detection are easy to occur in a high-density and high-speed scene, and reliable real-time tracking is difficult to realize. In a first aspect, an embodiment of the present application provides a method for determining a droplet flow rate, including: Acquiring a continuous image sequence of the liquid drops; Inputting the continuous image sequence to a trained target detection model for processing to obtain the droplet information of each droplet in each frame of droplet image in the continuous image sequence; determining a cross-frame associated droplet combination based on droplet information of each droplet in each frame of droplet images, wherein the droplet combination comprises a first droplet and a second droplet which are the same droplet in any two frames of adjacent images; and calculating the speed information of the liquid drops based on the liquid drop information of the liquid drop combination. Optionally, the determining a droplet combination associated across frames based on the droplet information of each droplet in each frame of droplet images includes: Acquiring adjacent frame images adjacent to the image where the first liquid drop is located in the continuous image sequence; Determining a plurality of candidate droplets associated with the first droplet in the adjacent frame image in a cross-frame manner based on the droplet information of the first droplet, the droplet information of each initial droplet in the adjacent frame image, and a candidate matching rule; determining the second droplet based on the droplet information of the first droplet, the droplet information of the plurality of candidate droplets, and a cross-frame correlation rule to obtain the droplet combination. Optionally, the droplet information includes droplet coordinate information and droplet size information; the determining a plurality of candidate droplets associated with the first droplet across frames in the adjacent frame image based on the droplet information of the first droplet, the droplet information of each initial droplet in the adjacent frame image, and a candidate matching rule includes: Calculating to obtain a coordinate distance between the first droplet and each initial droplet based on the droplet coordinate information of the first droplet and the droplet coordinate information of each initial droplet; Calculating a size change ratio between the first droplet and each initial droplet based on the droplet size information of the first droplet and the droplet size information of each initial droplet; And aiming at any initial liquid drop, if the coordinate distance corresponding to the initial liquid drop is smaller than a set distance and the size change proportion corresponding to the initial liquid drop is smaller than a set proportion, determining the initial liquid drop as the candidate liquid drop. Optionally, the determining the second droplet based on the droplet information of the first droplet, the droplet information of the plurality of candidate droplets, and the cross-frame association rule includes: determining a target candidate liquid drop with the minimum coordinate distance and the minimum size change proportion from a plurality of candidate liquid drops; determining a flow direction between the first droplet and the target candidate droplet based on the droplet coordinate information of the first droplet and the droplet coordinate information of the target candidate droplet; and if the flow