Automatic crop health change detection and alerting system
地域：
Winnipeg
摘要
A method and system for crop health change monitoring is provided. The method includes acquiring a companion image of a crop growing within a field at a first point in time, acquiring a master image of the crop growing within the field at a second point in time, and computing, using a processor, vegetation indices using the master image and the companion image, determining, using the processor, regions of change within the master image using the vegetation indices and generating an alert indicative of a change in crop condition of the crop growing within the field, and communicating the alert indicative of the change in crop condition over a network to a computing device configured to receive the alert.
說明書
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
This application claims priority to U.S. Provisional Application No. 62/692,416, filed Jun. 29, 2018, hereby incorporated by reference in its entirety.
This description relates to the detection of changes in crop health condition within an agricultural field. More specifically, this description relates to the use of remotely-sensed image data for automatic detection of regions of change within the field.
Remotely-sensed image data and products derived from that data (i.e., imagery products) are increasingly utilized in agriculture. This is because these data products can provide rapid, synoptic estimates of crop health condition over a large number of agricultural acres. Crop health condition can be estimated using vegetation indices derived from the original image spectral data. One example vegetation index is the Normalized Difference Vegetation Index (NDVI), which can demonstrate high correlations with crop biomass, productivity, and eventual yield. NDVI and other imagery products can also provide quantitative and visual indications of deleterious crop conditions such as pest, disease, or weather damage (i.e., hail), as well as the presence of weeds.
Despite the utility offered by these imagery products, manual inspection of images can be very time consuming and tedious. This can be particularly true for growers operating very large farming operations. Manual inspection of images and imagery products can also require expertise and experience to properly interpret the data. As such, a method to automatically detect and highlight potential crop issues is desirable.
權(quán)利要求
What is claimed is:1. A method for crop health change monitoring, the method comprising:(i) acquiring a companion image of a crop growing within a field at a first point in time; (ii) acquiring a master image of the crop growing within the field at a second point in time, the first point in time prior to the second point in time; (iii) computing, using a processor, vegetation indices using the master image and the companion image; (iv) determining, using the processor, regions of change within the master image using the vegetation indices by:(a) subtracting a companion normalized vegetation index image from a master normalized vegetation index image based on the master image to create a change image; (b) flagging change pixels in the change image exceeding a change threshold; and (c) removing contiguous groupings of flagged change pixels in the change image responsive to the contiguous groupings being smaller than a region size threshold; (v) generating an alert indicative of a change in crop condition of the crop growing within the field between the first point in time and the second point in time responsive to change within one or more of the regions of change being sufficient to meet defined criteria; and (vi) communicating the alert indicative of the change in crop condition over a network to a computing device configured to receive the alert. 2. The method of claim 1 further comprising selecting the companion image from a stack of images based on at least one candidate selection criteria. 3. The method of claim 2 wherein the at least one candidate selection criteria comprises a date range parameter and at least one of a minimum threshold for an average vegetation index value in the field, a maximum threshold for an average vegetation index value in the field, and a growth stage parameter. 4. The method of claim 1 wherein the defined criteria comprises a minimum threshold. 5. The method of claim 1 wherein the alert comprises a change alert image indicative of the change in crop condition of the crop growing within the field between the first point in time and the second point in time. 6. The method of claim 1 further comprising removing flagged change pixels if a master image normalized vegetation index value associated with the master normalized index image base is over a specified master image normalized vegetation index threshold. 7. The method of claim 1 further comprising removing flagged change pixels if a master image normalized vegetation index value associated with the master normalized index image base is under a specified master image normalized vegetation index threshold. 8. The method of claim 1 further comprising acquiring the companion image by:determining companion image candidates; computing class area changes between the master image and the companion image candidates and applying change thresholds to the class area changes to determine one or more eligible images; and selecting the companion image from the one or more eligible images using candidate image selection criteria. 9. The method of claim 1 further comprising comparing the one or more of the regions of change with one or more regions of change of a previously generated alert to determine one or more regions of new change, the step of generating the alert responsive to change within one or more of the regions of change being sufficient to meet the defined criteria further comprising applying a minimum cumulative change area threshold to the one or more regions of new change. 10. The method of claim 1 further comprising applying image filtering to the master image and the companion image prior to computing the vegetation indices. 11. The method of claim 1 wherein the vegetation indices are normalized difference vegetation indices. 12. A method for crop health change monitoring, the method comprising:(i) acquiring a companion image of a crop growing within a field at a first point in time, the companion image being acquired by:(a) determining companion image candidates; (b) computing class area changes between the master image and the companion image candidates and applying change thresholds to the class area changes to determine one or more eligible images; and (c) selecting the companion image from the one or more eligible images using candidate image selection criteria; (ii) acquiring a master image of the crop growing within the field at a second point in time, the first point in time prior to the second point in time; (iii) computing, using a processor, vegetation indices using the master image and the companion image; (iv) determining, using the processor, regions of change within the master image using the vegetation indices; (v) generating an alert indicative of a change in crop condition of the crop growing within the field between the first point in time and the second point in time responsive to change within one or more of the regions of change being sufficient to meet defined criteria; and (vi) communicating the alert indicative of the change in crop condition over a network to a computing device configured to receive the alert. 13. The method according to claim 12 wherein the step of selecting the companion image from the one or more eligible images using candidate image selection criteria further comprises:computing for each of the one or more eligible images an excess value representative of an amount that the class area changes exceed the change thresholds; and selecting the eligible image having the excess value which is greatest. 14. A method for crop health change monitoring, the method comprising:(i) acquiring a companion image of a crop growing within a field at a first point in time; (ii) acquiring a master image of the crop growing within the field at a second point in time, the first point in time prior to the second point in time; (iii) computing, using a processor, vegetation indices using the master image and the companion image; (iv) determining, using the processor, regions of change within the master image using the vegetation indices; (v) generating an alert indicative of a change in crop condition of the crop growing within the field between the first point in time and the second point in time responsive to:(a) change within one or more of the regions of change being sufficient to meet defined criteria; and (b) subsequent to comparing the one or more of the regions of change with one or more regions of change of a previously generated alert to determine one or more regions of new change, change within the one or more of the regions of new change being sufficient to meet a minimum cumulative change area threshold; and (vi) communicating the alert indicative of the change in crop condition over a network to a computing device configured to receive the alert.
意見反饋
使用該功能遇到問題