Intelligent weed treatment

1 . A method comprising:

accessing, by a farming machine, an image of a portion of a geographic area;

analyzing, by the farming machine, the image to identify a plant in the portion of the geographic area;

accessing, by the farming machine, a set of treatment objectives, the set of treatment objectives including two or more of a plant hindrance objective, a plant reproduction limiting objective, a carbon footprint limiting objective, and a collateral damage limiting objective;

determining, by the farming machine, a regime zone corresponding to the portion of the geographic area;

accessing, for each of a plurality of treatment actions, treatment score data indicating efficacy of the treatment action in achieving a given treatment objective for a given type of plant;

accessing regime zone data indicating a subset of one or more of the plurality of treatment actions that are impermissible for the regime zone; and

determining, by the farming machine, one or more treatment actions of a treatment plan for the plant based on one or more characteristics of the plant, the set of treatment objectives, and the regime zone corresponding to the portion of the geographic area, wherein the one or more treatment actions are selected from among the plurality of treatment actions by excluding the subset of treatment actions and selecting, from a remainder of the plurality of treatment actions, the one or more treatment actions based on the treatment score data.

2 . The method of claim 1 , wherein analyzing the image comprises segmenting the image into regions corresponding to identified one or more plant components of the plant, and

wherein the method further comprises determining the one or more characteristics of the plant based on the identified one or more plant components.

3 . The method of claim 2 , wherein the one or more treatment actions include at least two treatment actions, and wherein the method further comprises:

performing a first treatment action for a first one of the one or more plant components of the plant, wherein the first treatment action is performed to address a first one of the set of treatment objectives; and

performing a second treatment action for a second one of the one or more plant components of the plant, wherein the second treatment action is performed to address a second one of the set of treatment objectives.

4 . The method of claim 1 , further comprising:

generating a georeferenced weed record based on the treatment plan;

transmitting the georeferenced weed record to an external device; or

performing the one or more treatment actions on the plant based on the treatment plan.

5 . The method of claim 4 , wherein the one or more treatment actions includes one or more of a mechanical treatment action, a chemical treatment action, a hot fluid treatment action, a pneumatic treatment action, a vacuum treatment action, a pressurized fluid treatment action, an electromagnetic radiation treatment action, or a biological treatment action.

6 . The method of claim 1 , wherein the plant is a weed and wherein the one or more characteristics of the plant include a type of the weed and a growth stage of the weed.

7 . The method of claim 6 , further comprising:

accessing the treatment score data based on the type of the weed, the treatment score data including treatment scores for combinations of treatment actions and treatment objectives; and

selecting the one or more treatment actions based on the treatment scores and the set of treatment objectives, wherein the treatment plan is determined based on the selected one or more treatment actions.

8 . The method of claim 1 , wherein determining the treatment plan comprises inputting the one or more characteristics and the set of treatment objectives into a machine-learned model that is trained based on historical treatment plan data indicating treatment actions performed for plants having different characteristics to output recommended one or more treatment actions.

9 . A farming machine comprising:

a hardware processor; and

a non-transitory computer-readable storage medium storing executable instructions that, when executed by the hardware processor, cause the hardware processor to perform steps comprising:

accessing an image of a portion of a geographic area;

analyzing the image to identify a plant in the portion of the geographic area;

accessing a set of treatment objectives, the set of treatment objectives including two or more of a plant hindrance objective, a plant reproduction limiting objective, a carbon footprint limiting objective, and a collateral damage limiting objective;

determining a regime zone corresponding to the portion of the geographic area;

accessing, for each of a plurality of treatment actions, treatment score data indicating efficacy of the treatment action in achieving a given treatment objective for a given type of plant;

accessing regime zone data indicating a subset of one or more of the plurality of treatment actions that are impermissible for the regime zone; and

determining one or more treatment actions of a treatment plan for the plant based on one or more characteristics of the plant, the set of treatment objectives, and the regime zone corresponding to the portion of the geographic area, wherein the one or more treatment actions are selected from among the plurality of treatment actions by excluding the subset of treatment actions and selecting, from a remainder of the plurality of treatment actions, the one or more treatment actions based on the treatment score data.

10 . The farming machine of claim 9 , wherein the instructions that cause the hardware processor to perform the step comprising analyzing the image comprise instructions that, when executed by the hardware processor, cause the hardware processor to perform a step comprising segmenting the image into regions corresponding to identified one or more plant components of the plant, and

wherein the instructions further cause the hardware processor to perform a step comprising determining the one or more characteristics of the plant based on the identified one or more plant components.

11 . The farming machine of claim 10 , wherein the one or more treatment actions include at least two treatment actions, and wherein the instructions further cause the hardware processor to perform steps comprising:

performing a first treatment action for a first one of the one or more plant components of the plant, wherein the first treatment action is performed to address a first one of the set of treatment objectives; and

performing a second treatment action for a second one of the one or more plant components of the plant, wherein the second treatment action is performed to address a second one of the set of treatment objectives.

12 . The farming machine of claim 9 , wherein the instructions further cause the hardware processor to perform steps comprising one or more of:

generating a georeferenced weed record based on the treatment plan;

transmitting the georeferenced weed record to an external device; or

performing the one or more treatment actions on the plant based on the treatment plan.

13 . The farming machine of claim 12 , wherein the one or more treatment actions includes one or more of a mechanical treatment action, a chemical treatment action, a hot fluid treatment action, a pneumatic treatment action, a vacuum treatment action, a pressurized fluid treatment action, an electromagnetic radiation treatment action, or a biological treatment action.

14 . The farming machine of claim 9 , wherein the plant is a weed and wherein the one or more characteristics of the plant include a type of the weed and a growth stage of the weed.

15 . The farming machine of claim 14 , wherein the instructions further cause the hardware processor to perform steps comprising:

accessing the treatment score data based on the type of the weed, the treatment score data including treatment scores for combinations of treatment actions and treatment objectives; and

selecting the one or more treatment actions based on the treatment scores and the set of treatment objectives, wherein the treatment plan is determined based on the selected one or more treatment actions.

16 . A non-transitory computer-readable storage medium storing executable instructions that, when executed by a hardware processor, cause the hardware processor to perform steps comprising:

accessing an image of a portion of a geographic area;

analyzing the image to identify a plant in the portion of the geographic area;

accessing a set of treatment objectives, the set of treatment objectives including two or more of a plant hindrance objective, a plant reproduction limiting objective, a carbon footprint limiting objective, and a collateral damage limiting objective;

determining a regime zone corresponding to the portion of the geographic area;

accessing, for each of a plurality of treatment actions, treatment score data indicating efficacy of the treatment action in achieving a given treatment objective for a given type of plant;

accessing regime zone data indicating a subset of one or more of the plurality of treatment actions that are impermissible for the regime zone; and

determining one or more treatment actions of a treatment plan for the plant based on one or more characteristics of the plant, the set of treatment objectives, and the regime zone corresponding to the portion of the geographic area, wherein the one or more treatment actions are selected from among the plurality of treatment actions by excluding the subset of treatment actions and selecting, from a remainder of the plurality of treatment actions, the one or more treatment actions based on the treatment score data.