Predictive models for visually classifying insects
AssigneeVerily Life Sciences LLC
InventorsTiantian Zha, Yaniv Ovadia, Daniel Newburger, Dilip Krishnan, Josh Livni, Mark Desnoyer
Life of this patent
Estimating the term…
Abstract
Insects can be localized and classified using a predictive model. To begin, image data is obtained that corresponds to the insects. Using a predictive model, samples of the image data are evaluated to determine whether the image portions include an insect and, if so, into what category the insect should be classified (e.g., male/female, species A/species B, etc.).
Claims30
A computer-implemented method, comprising:
- accessing an image depicting a plurality of insects on a surface;
- determining, using a predictive model, locations of individual insects of the plurality of insects in the image;
- classifying, using the predictive model, the individual insects into at least one category based at least in part on an insect parameter;
- and
- outputting insect information from the predictive model based at least in part on the locations of the individual insects and the at least one category.
A system, comprising:
- an imaging device configured to capture images of insects supported by a surface;
- and
- a computing device in communication with the imaging device, and configured to at least:
- cause the imaging device to capture an image of the surface, the image depicting a plurality of insects;
- determine, using a predictive model, locations of individual insects of the plurality of insects in the image;
- classify, using the predictive model, the individual insects into at least one category based at least in part on an insect characteristic;
- determine whether additional images should be captured based on a confidence value;
- and
- output insect information from the predictive model based at least in part on the locations of the individual insects and the at least one category.
One or more non-transitory computer-readable medium comprising computer-executable instructions that, when executed by one or more computer systems, cause the computer systems to perform operations comprising:
- accessing an image depicting a plurality of insects;
- determining, using a predictive model, locations of individual insects of the plurality of insects in the image;
- classifying, using the predictive model, the individual insects into at least one category based at least in part on a mosquito characteristic;
- and
- outputting insect information from the predictive model based at least in part on the locations of the individual insects and the at least one category.
A computer-implemented method, comprising:
- accessing image data representing at least one mosquito on a surface;
- evaluating, using a predictive model, a first image portion of the plurality of image portions by at least:
- (i) locating the at least one mosquito represented in the first image portion;
- and
- (ii) classifying the at least one mosquito into at least one category based at least in part on an insect parameter;
- and
- outputting insect information from the predictive model based at least in part on locating the at least one mosquito and classifying the at least one mosquito.
Dependencies are inferred from the claim text.