Enhanced object detection
Methods, systems, and apparatus, including computer programs encoded on computer-storage media, for monocular underwater camera biomass estimation. In some implementations, an exemplary method includes obtaining an image of a fish captured by an underwater camera; identifying portions of the image corresponding to one or more areas of interest; extracting the portions of the image from the image; providing the portions of the image to a model trained to detect objects in the portions of the image; and determining an action based on output of the model indicating a number of object detections.
1. A method comprising:
obtaining an image of a fish captured by an underwater camera;
identifying portions of the image corresponding to one or more areas of interest;
extracting the portions of the image from the image;
determining, at different times, a processing load of a model that is trained to detect objects in the portions of the image;
when the processing load of the model satisfies a threshold at an earlier time, storing the extracted portions of the image in a database of portions that are to be later processed by the model;
when the processing load of the model does not satisfy the threshold at a later time, providing the stored portions of the image to the model; and
determining an action based on output of the model indicating a number of object detections.
2. The method of claim 1 , comprising:
obtaining a type of object to be detected by the trained model.
3. The method of claim 2 , comprising:
determining one or more of the areas of interest as areas that typically include the type of object to be detected.
4. The method of claim 1 , wherein the objects include parasites.
5. The method of claim 4 , wherein the areas of interest include regions behind dorsal or adipose fins.
6. The method of claim 1 , comprising:
generating a value indicating a level of infestation within a population that includes the fish.
7. The method of claim 6 , wherein the value is generated for the population by a model trained to determine infestation for a given population using objects detected on a portion of the given population.
8. The method of claim 1 , comprising:
detecting the fish within the image using a model trained to detect fish.
9. The method of claim 8 , wherein the areas of interest include portions of the detected fish.
10. The method of claim 1 , wherein the action comprises:
adjusting a feeding system providing feed to the fish.
11. The method of claim 1 , wherein the action comprises:
sending data indicating the output of the model to a user device, wherein the data is configured to, when displayed on the user device, present a user of the user device with a visual representation of disease in a population that includes the fish.
12. The method of claim 1 , wherein determining the action comprises:
determining to adjust a position or operation of an item of motorized equipment.
13. A non-transitory, computer-readable medium storing one or more instructions executable by a computer system to perform operations comprising:
obtaining an image of a fish captured by an underwater camera;
identifying portions of the image corresponding to one or more areas of interest;
extracting the portions of the image from the image;
determining, at different times, a processing load of a model that is trained to detect objects in the portions of the image;
when the processing load of the model satisfies a threshold at an earlier time, storing the extracted portions of the image in a database of portions that are to be later processed by the model;
when the processing load of the model does not satisfy the threshold at a later time, providing the stored portions of the image to the model; and
determining an action based on output of the model indicating a number of object detections.
14. The non-transitory, computer-readable medium of claim 13 , wherein the operations comprise:
obtaining a type of object to be detected by the trained model.
15. The non-transitory, computer-readable medium of claim 14 , wherein the operations comprise:
determining one or more of the areas of interest as areas that typically include the type of object to be detected.
16. A computer-implemented system, comprising:
one or more computers; and
one or more computer memory devices interoperably coupled with the one or more computers and having tangible, non-transitory, machine-readable media storing one or more instructions that, when executed by the one or more computers, perform one or more operations comprising:
obtaining an image of a fish captured by an underwater camera;
identifying portions of the image corresponding to one or more areas of interest; extracting the portions of the image from the image;
determining, at different times, a processing load of a model that is trained to detect objects in the portions of the image;
when the processing load of the model satisfies a threshold at an earlier time, storing the extracted portions of the image in a database of portions that are to be later processed by the model;
when the processing load of the model does not satisfy the threshold at a later time, providing the stored portions of the image to the model; and
determining an action based on output of the model indicating a number of object detections.