Variable rate sprayer system and method of variably applying agrochemicals
Systems, methods and computer-readable media are provided for controlling a variable-rate sprayer for precision agriculture. Highly efficient digital image processing enables rapid and reliable control of the variable rate sprayer. In one embodiment, image processing uses only a subset of luminance, hue, saturation and intensity textural features to provide rapid image recognition. In another embodiment, an image is decomposed into RGB components and a G is ratio determined. For example, the textural method is useful in growing season where color differentiation is difficult. The G ratio method is useful in early spring and late fall where color differentiation is possible. These rapid computationally light methods enable a mobile sprayer system to identify crop or field conditions in real-time and to dispense an appropriate amount of agrochemical in a specific section of the sprayer boom where the target has been detected as the mobile sprayer advances.
1. A mobile variable rate sprayer system for applying an agrochemical, the system comprising:
a vehicle for transporting the agrochemical;
at least one digital camera mounted to the vehicle for capturing an image;
a processor for processing the image to determine image features using only a subset of luminance, hue, saturation and intensity image features;
at least one controller for receiving signals from the processor; and
at least one variable rate sprayer mounted on the vehicle, wherein each spray nozzle of the sprayer is controlled by the controller based on the image features.
2. The system as claimed in claim 1 wherein the vehicle comprises a front boom for supporting a plurality of spaced-apart digital cameras and a rear boom for supporting a plurality of spaced-apart spray nozzles and wherein the controller is a multi-channel controller for controlling the plurality of spray nozzles.
3. The system as claimed in claim 1 wherein the processor is configured to use as the subset only textural features drawn from hue, saturation and intensity co-occurrence matrices.
4. The system as claimed in claim 1 wherein the processor is configured to use as the subset only textural features drawn from saturation and intensity co-occurrence matrices.
5. A method of variably applying an agrochemical, the method comprising:
capturing an image;
processing the image to determine image features by using only a subset of luminance, hue, saturation and intensity image features; and
controlling a variable rate spray nozzle based on the image features to thereby deliver an appropriate amount of the agrochemical.
6. The method as claimed in claim 5 wherein processing the image comprises:
pre-processing the image by converting the image into NTSC-standard luminance, hue, saturation and intensity images;
dividing each image into sub-regions;
composing co-occurrence matrices for each of the sub-regions; and
normalizing matrices to extract the textural features.
7. The method as claimed in claim 6 wherein the subset comprises only textural features drawn from hue, saturation and intensity co-occurrence matrices.
8. The method as claimed in claim 6 wherein the subset comprises only textural features drawn from saturation and intensity co-occurrence matrices.
9. A non-transitory computer-readable medium comprising instructions in code which when executed by a processor of a computing device cause the computing device to:
send a signal to digital camera to cause the digital camera to capture an image;
receive image data from the digital camera;
process the image to determine image features by using only a subset of luminance, hue, saturation and intensity image features; and
control an agricultural device based on the image features.
10. The non-transitory computer-readable medium as claimed in claim 9 wherein the code is configured to process the image by:
pre-processing the image by converting the image into NTSC-standard luminance, hue, saturation and intensity images;
dividing each image into sub-regions;
composing co-occurrence matrices for each of the sub-regions; and
normalizing matrices to extract the textural features.
11. The non-transitory computer-readable medium as claimed in claim 10 wherein the subset comprises only textural features drawn from hue, saturation and intensity co-occurrence matrices.
12. The non-transitory computer-readable medium as claimed in claim 10 wherein the subset comprises only textural features drawn from saturation and intensity co-occurrence matrices.
13. The non-transitory computer-readable medium as claimed in claim 9 wherein controlling the agricultural device comprises controlling a variable rate spray nozzle based on the image features to thereby deliver an appropriate amount of the agrochemical.
14. A mobile variable rate sprayer system for applying an agrochemical, the system comprising:
a vehicle for transporting the agrochemical;
at least one digital camera mounted to the vehicle for capturing an image;
a processor for processing the image into RGB components and to determine a G ratio by computing a ratio of G to the sum of R, G and B;
at least one controller for receiving signals from the processor; and
at least one variable rate sprayer mounted on the vehicle, wherein each spray nozzle of the sprayer is controlled by the controller based on the G ratio.
15. The system as claimed in claim 14 wherein the vehicle comprises a front boom for supporting a plurality of spaced-apart digital cameras and a rear boom for supporting a plurality of spaced-apart spray nozzles and wherein the controller is a multi-channel controller for controlling the plurality of spray nozzles.
16. A method of variably applying an agrochemical, the method comprising:
capturing an image;
processing the image into RGB components and to determine a G ratio by computing a ratio of G to the sum of R, G and B; and
controlling a variable rate spray nozzle based on the G ratio to thereby deliver an appropriate amount of the agrochemical.
17. A non-transitory computer-readable medium comprising instructions in code which when executed by a processor of a computing device cause the computing device to:
send a signal to digital camera to cause the digital camera to capture an image;
receive image data from the digital camera;
process the image data into RGB components and to determine a G ratio by computing a ratio of G to the sum of R, G and B;
control an agricultural device based on the G ratio.
18. The non-transitory computer-readable medium as claimed in claim 17 wherein the code is configured to control a variable rate spray nozzle based on the G ratio to thereby deliver an appropriate amount of the agrochemical.