System and method for classification of particles in a fluid sample
First Claim
Patent Images
1. A system for classifying particles in water sample, system comprising:
- an optical detection assembly comprising at least one image acquisition device and corresponding lens with an optical axis, the image acquisition device is configured to acquire images of an image acquisition area perpendicular to the optical axis;
a sample device comprising at least one sample container suitable for holding a water sample in liquid form comprising particles;
a translating arrangement configured to translate said image acquisition area through at least a part of said sample container with an optimal step length between each acquired image, wherein the optimal step length is determined based on a depth of field of the optical detection assembly, and a selected optical resolution;
a controller configured to control said optical detection assembly and said translating arrangement to acquire images of a plurality of image acquisition areas;
an image processor programmed to analyze said acquired images and to determine a set of features for an individual particle captured by said acquired images by creating objects of the individual particle and assigning each of the objects to a corresponding stack of objects comprising different images of the individual particle; and
an artificial intelligent processor programmed to associate said set of features for said individual particle to a particle classification of particles,wherein said step length between each acquired image is used to determine a distance between the objects in the stacks of objects.
0 Assignments
0 Petitions
Accused Products
Abstract
A particle classifier system and a method of training the system are described. The particle classifier system is suitable for classification of particles in a liquid sample, said particle classifier system comprises
- an optical detection assembly comprising at least one image acquisition device with an optical axis, the image acquisition device is configured to acquire images of an image acquisition area perpendicular to said optical axis;
- a sample device comprising at least one sample container suitable for holding a sample in liquid form;
- a translating arrangement configured to translate said image acquisition area through at least a part of said sample container;
- a control system configured to controlling said optical detection assembly and said translating unit to acquire images of a plurality of image acquisition areas;
- an image analyzing processing system programmed to analyze said acquired images by a method comprising
creating objects (sub-images) of individual particles captured by said acquired images,
creating stacks of objects of each individual particle,
identifying complete stacks of objects comprising
at least one object wherein said particle is in-focus, and
two objects wherein said particle is out-of-focus, and
determining, for each of said complete stacks of objects, a set of values for a set of features of at least N features, wherein N is larger than or equal to 1, and wherein the determination of said values of said set of features involve data obtained from said at least one object wherein said particle is in-focus, and/or said at least two objects wherein said particle is out-of-focus; and - an artificial intelligent processing system programmed to associate said set of values for said determined set of features for each individual particle to a particle classification.
-
Citations
14 Claims
-
1. A system for classifying particles in water sample, system comprising:
-
an optical detection assembly comprising at least one image acquisition device and corresponding lens with an optical axis, the image acquisition device is configured to acquire images of an image acquisition area perpendicular to the optical axis; a sample device comprising at least one sample container suitable for holding a water sample in liquid form comprising particles; a translating arrangement configured to translate said image acquisition area through at least a part of said sample container with an optimal step length between each acquired image, wherein the optimal step length is determined based on a depth of field of the optical detection assembly, and a selected optical resolution; a controller configured to control said optical detection assembly and said translating arrangement to acquire images of a plurality of image acquisition areas; an image processor programmed to analyze said acquired images and to determine a set of features for an individual particle captured by said acquired images by creating objects of the individual particle and assigning each of the objects to a corresponding stack of objects comprising different images of the individual particle; and an artificial intelligent processor programmed to associate said set of features for said individual particle to a particle classification of particles, wherein said step length between each acquired image is used to determine a distance between the objects in the stacks of objects. - View Dependent Claims (2, 3, 4, 7, 8, 9, 10, 11, 12)
-
-
5. A method of creating objects of individual particles in a water sample held in a sample container, the method comprising:
-
determining an optimal step length based on a depth of field of the optical detection assembly, and a selected optical resolution; translating an image acquisition area, formed by an image acquisition device and corresponding lens in the sample container, to multiple positions in the sample container, the image acquisition area being perpendicular to an optical axis of the lens; acquiring images of the image acquisition area at the multiple positions in the sample container; determining a set of features for an individual particle in the water sample captured by the acquired images by creating objects of the individual particle, wherein creating the objects of the individual particle comprises; copying all pixels within an object periphery line surrounding the individual particle; and subjecting said copied pixels to a filtration in at least one filter comprising a shape-filter, a size-filter, a contrast-filter or an intensity filter. - View Dependent Claims (6)
-
-
13. A method of analyzing a test water sample held in a sample container for the presence of a target particle, the method comprising:
-
providing the test water sample; determining an optimal step length based on a depth of field of the optical detection assembly, and a selected optical resolution; translating an image acquisition area, formed by an image acquisition device and corresponding lens in the sample container, to multiple positions in the sample container; acquiring at least 3 images of individual particles, respectively, in said test water sample at multiple positions in the sample container, wherein said individual particles respectively are in-focus or out-of-focus; creating objects of said individual particles, respectively, captured by said acquired images and providing stacks of objects for each individual particle; identifying complete stacks of objects, each identified complete stack of objects at least comprising an object wherein the corresponding individual particle is in-focus and two objects wherein the corresponding individual particle is out-of-focus; determining a set of features of at least N features for each of said identified complete stacks of objects, wherein N is larger than or equal to 1; and associating said set of values for said determined set of features for each individual particle to classify said individual particle and to determine if at least one target particle has been classified from among the individual particles. - View Dependent Claims (14)
-
Specification