Patent Application
20180018414
This U.S. patent application claims priority under 35 U.S.C. §119 to: India Application No. 201621023796 filed on Jul. 12, 2016. The entire contents of the aforementioned application are incorporated herein by reference.
The embodiments herein generally relate to pest management, and more particularly to systems and methods for pest forecasting using historical pesticide usage information.
Agriculture plays an important role in sustaining and enhancing human life. Accordingly, a lot of research is prevalent in ecological forecasting in agricultural domain for forecasting of pest, disease infestation instances, pest severity and disease severity. Traditionally, surveillance data collected on pest or disease infestation instances are relied upon for forecasting model generation or training of the models. As such, voluminous data of pest or disease incidences are required to build a reliable model. Collecting these data through surveillance method is tedious, time consuming and costly as it also requires immense manpower to collect the data from fields. Also, researchers need to wait for a long time period to generate a good training dataset for reliable model generation. In case of field or site specific forecasting model generation it becomes very much difficult to collect such huge amount of data from each of the fields.
Embodiments of the present disclosure present technological improvements as solutions to one or more of the above-mentioned technical problems recognized by the inventors in conventional systems.
Systems and methods of the present disclosure facilitate generating field or site specific pest forecasting models in an efficient and time and cost effective manner.
In an aspect, there is provided a processor implemented method comprising: receiving a first set of inputs pertaining to pesticide application associated with a pest in a geo-location under consideration and agronomic information thereof; estimating, based on the first set of inputs, a time instance (t0) pertaining to an affecting stage where the pest starts affecting vegetation in the geo-location; and estimating a set of time instances (t−1) pertaining to an inoculation stage based on the time instance (t0) and a pre-defined range of weather conducive lag (τ) between each time instance (t−1) from the set of time instances (t−1) and the time instance (t0), the pre-defined range of weather conducive lag (τ) being characterized by life cycle of the pest or disease affecting the vegetation.
In another aspect, there is provided a system comprising: one or more data storage devices operatively coupled to the one or more processors and configured to store instructions configured for execution by the one or more processors to: receive a first set of inputs pertaining to pesticide application associated with a pest in a geo-location under consideration and agronomic information thereof; estimate, based on the first set of inputs, a time instance (t0) pertaining to an affecting stage where the pest starts affecting vegetation in the geo-location; and estimate a set of time instances (t−1) pertaining to an inoculation stage based on the time instance (t0) and a pre-defined range of weather conducive lag (τ) between each time instance (t−1) from the set of time instances (t−1) and the time instance (t0), the pre-defined range of weather conducive lag (τ) being characterized by life cycle of the pest or disease affecting the vegetation.
In yet another aspect, there is provided a computer program product comprising a non-transitory computer readable medium having a computer readable program embodied therein, wherein the computer readable program, when executed on a computing device, causes the computing device to: receive a first set of inputs pertaining to pesticide application associated with a pest in a geo-location under consideration and agronomic information thereof; estimate, based on the first set of inputs, a time instance (t0) pertaining to an affecting stage where the pest starts affecting vegetation in the geo-location; and estimate a set of time instances (t−1) pertaining to an inoculation stage based on the time instance (t0) and a pre-defined range of weather conducive lag (τ) between each time instance (t−1) from the set of time instances (t−1) and the time instance (t0), the pre-defined range of weather conducive lag (τ) being characterized by life cycle of the pest or disease affecting the vegetation.
In an embodiment of the present disclosure, the one or more hardware processors are further configured to receive a second set of inputs pertaining to historical weather data associated with a geo-location under consideration; and generate a pest forecasting model for each pest corresponding to each time instance (t−1) from the set of time instances pertaining to the inoculation stage (t−1), the pest forecasting model being trained based on at least a portion of data received over a pre-defined time period, pertaining to (i) the time instance (t0) (ii) the time instances (t−1) and (iii) the second set of inputs.
In an embodiment of the present disclosure, the one or more hardware processors are further configured to test, for accuracy, the pest forecasting model corresponding to each time instance (t−1) based on remaining portion of the data received over the pre-defined time period, pertaining to (i) the time instance (t0) (ii) the time instances (t−1), and (iii) the second set of inputs over the pre-defined time period; and generate an optimum pest forecasting model based on an optimum weather conducive lag from the pre-defined range of weather conducive lag (τ), the optimum weather conducive lag generating highest accuracy.
In an embodiment of the present disclosure, the one or more hardware processors are further configured to validate and adapt the optimum pest forecasting model based on reported first set of inputs.
In an embodiment of the present disclosure, the reported first set of inputs comprises one or more of participatory sensing inputs and crowd sourcing inputs pertaining to the agronomic information.
In an embodiment of the present disclosure, the one or more processors are further configured to validate and adapt the optimum pest forecasting model by: receiving images from the reported first set of inputs; filtering the received images to identify pest/disease patches and extract corresponding portion of vegetation as region of interest (ROI) in the images; analyzing one or more features, in the ROI, from the group of features consisting of color, shape and texture; computing maturity of pest based on the analyzed one or more features; and adapting the optimum forecasting model based on the computed maturity of pest.
It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the embodiments of the present disclosure, as claimed.
The embodiments herein will be better understood from the following detailed description with reference to the drawings, in which:
It should be appreciated by those skilled in the art that any block diagram herein represent conceptual views of illustrative systems embodying the principles of the present subject matter. Similarly, it will be appreciated that any flow charts, flow diagrams, state transition diagrams, pseudo code, and the like represent various processes which may be substantially represented in computer readable medium and so executed by a computing device or processor, whether or not such computing device or processor is explicitly shown.
Exemplary embodiments are described with reference to the accompanying drawings. In the figures, the left-most digit(s) of a reference number identifies the figure in which the reference number first appears. Wherever convenient, the same reference numbers are used throughout the drawings to refer to the same or like parts. While examples and features of disclosed principles are described herein, modifications, adaptations, and other implementations are possible without departing from the spirit and scope of the disclosed embodiments. It is intended that the following detailed description be considered as exemplary only, with the true scope and spirit being indicated by the following claims.
Before setting forth the detailed explanation, it is noted that all of the discussion below, regardless of the particular implementation being described, is exemplary in nature, rather than limiting.
Generation of ecological models like pest or disease forecasting models requires ground truth on actual incidences of pest or diseases in the field. Collection of information on pest or disease incidence instances using surveillance method is very time consuming and costly. Most farmers or agricultural organizations maintain data pertaining to usage of chemicals/pesticides in their farms for cost-benefit analysis which the present disclosure utilizes for development of forecasting models In accordance with the present disclosure, there is no added effort required to be taken by farmers for such data collection and data from large number of farms at a given instance can be collected easily as compared to the traditional approach wherein visits to farms is required for collecting such data.
Thus, the present disclosure addresses the problem of providing field specific pest or disease forecasting models generated based on field specific farming practices. The present disclosure provides systems and methods for generating pest instance forecasting model based on historical information on pesticide usage and field specific farming practices or agronomic information that is easily available as compared to data pertaining to actual pest instances. Historical pesticide data like instance of pesticide application, quantity of the pesticide used, fraction of total section area covered at each instances of pesticide spray, type of spray (like spot spray or blanket spray) applied can give information about the actual pest infestation instances. In an embodiment, pest instance forecasting can lead to disease forecasting. Pest infestation instances can be detected correctly from pesticide data only when the pesticide usage is prophylactic in nature, i.e. pesticide has been applied in the farm only when the pest infestation in the farm has been identified. A typical method of preparing ground truth on pest infestation instances includes normalizing pesticide quantity/area covered within 1-6 days for each instance of pesticide application. Pest infestation instances is set to normalized value of pesticide quantity minus the pesticide application date. Pest infestation is then assumed to be sustained for next 5 days.
Referring now to the drawings, and more particularly to
The I/O interface device(s) 106 can include a variety of software and hardware interfaces, for example, a web interface, a graphical user interface, and the like and can facilitate multiple communications within a wide variety of networks N/W and protocol types, including wired networks, for example, LAN, cable, etc., and wireless networks, such as WLAN, cellular, or satellite. In an embodiment, the I/O interface device(s) can include one or more ports for connecting a number of devices to one another or to another server.
The memory 102 may include any computer-readable medium known in the art including, for example, volatile memory, such as static random access memory (SRAM) and dynamic random access memory (DRAM), and/or non-volatile memory, such as read only memory (ROM), erasable programmable ROM, flash memories, hard disks, optical disks, and magnetic tapes. In an embodiment, one or more modules (not shown) of the system 100 can be stored in the memory 102.
In an embodiment of the present disclosure, at step 202, the one or more processors 104 receive a first set of inputs pertaining to pesticide application associated with a pest in a geo-location under consideration and agronomic information pertaining to that geo-location. In an embodiment, pesticide inputs pertaining to pesticide application may include one or more of date of pesticide application, quantity of pesticide, dosage of pesticide, area covered during each instance of pesticide application, type of pesticide, method of application, and the like. Likewise, in an embodiment, agronomic information may include crop specific information, terrain and vegetation information, and the like that may be received in any form such as text or images. In an embodiment, agronomic information may be contextual data or crop specific data, data pertaining to methods used by farmers to raise crops and care for the soil, farming practices like pesticide application schedules, if pesticide is applied repeatedly or in periodic manner, and the like obtained directly from farmers or via satellites, UAVs, terrestrial sensors, and such other systems that operate on geo-spatial technologies and provide geo-spatial datasets.
In an embodiment of the present disclosure, at step 204, the one or more processors 104 estimate the time instance (t0) pertaining to an affecting stage where the pest starts affecting vegetation in the geo-location based on the received first set of inputs at step 202. In an embodiment, pseudo code for estimating the time instance (t0) may be represented as—
Identify pesticide used for a particular disease.
Initialize disease risk for each day in study period to low (0).
Keep flag=0 for each day to keep track whether pesticide application of that day has been already checked or not.
For each date tin study period
Variable effects of pesticides on pest prediction over different days lag has to be considered when developing the forecasting model. In an embodiment, system of the present disclosure computes the time period which can create a conducive condition for the pest to grow to that stage when it starts affecting the plant and causes the infestation in consecutive time instances. Weather of past few days from the instance of visual appearance of the pest or the stage of the pest when it starts affecting the plant will have strong correlations or effect on the consecutive pest instances. In this disclosure, this time period is defined as weather conducive lag. This weather conducive lag will vary from pest to pest and disease to disease based on their life cycle. The system 100 of the present disclosure supports identification of the weather conducive lag for different pests or diseases. Based on available information on pest life cycle or disease cycle, the weather conducive lag can be varied to identify the optimum weather conducive lag which will have highest correlation with the consecutive pest infestation instances.
In an embodiment of the present disclosure, at step 206, the one or more processors 104 estimate a set of time instances (t−1) pertaining to an inoculation stage based on the time instance (t0) and a pre-defined range of weather conducive lag (τ) between each time instance (t−1) from the set of time instances (t−1) and the time instance (t0) (206). In an embodiment, the pre-defined range of weather conducive lag (i) is characterized by life cycle of the pest or disease affecting the vegetation.
In an embodiment, pseudo code for estimating the time instance (t−1) may be represented as—
Define weather conducive lag (τ), as τ=(t0−t−1).
Identify the range to vary τ, based on information gathered from literature regarding the life cycle of the pest or disease.
For each lag, τ=6 to 30 (as per literature time span required by the pest to grow to larva or adult stage, value range indicated being specific to a pest ‘tea looper’)
Train a forecasting model such as Random Forest Model (from literature) using 80% of the samples.
Test developed forecasting model on rest 20% samples and find test set accuracy.
Assign weather conducive lag as T, for which test set accuracy is highest.
Final or optimum forecasting model is defined as the model which gives best test set accuracy.
Thus, in an embodiment, the one or more processors 104 generate an optimum pest forecasting model based on an optimum weather conducive lag from the pre-defined range of weather conducive lag (τ), wherein the optimum weather conducive lag generates highest accuracy.
Output: This technique returns a value of weather conducive lag τ. t−1 is defined as, t0−τ.
Weather information aids in identifying past pest infestation instances because pesticide usage is also dependent on some weather factors like rainfall. If after a pesticide spray there is a heavy rainfall then pesticide application is repeated in that case. So, to decide pest infestation instances from pesticide usage, these factors need to be considered. In an embodiment, the method 200 further provides receiving a second set of inputs pertaining to historical weather data associated with a geo-location under consideration. In an embodiment, the weather inputs may be received via Wireless Sensor Networks (WSN), Automatic Weather Stations (AWS), weather satellites, and the like. In an embodiment, the weather inputs may comprise data pertaining to temperature, rainfall, humidity, and such other weather parameters that are critical to crop management and pest control. In an embodiment, the second set of inputs may be statistical variability of weather parameters of past few days (varied within a range of 6-30) from each day in the study period.
In an embodiment, the one or more processors 104 generate a pest forecasting model for each pest corresponding to each time (t−1) from the set of time instances pertaining to the inoculation stage (t−1). The pest forecasting model is trained based on at least a portion of data received over a pre-defined time period, pertaining to (i) the time instance (to) (ii) the time instances (t−1) and (iii) the second set of inputs. In an embodiment, the at least a portion of data may be 80% of the received data. The one or more processors 104 further test the pest forecasting model corresponding to each time instance (t−1) for accuracy based on remaining portion of the data received over the pre-defined time period. In an embodiment, the remaining portion of the data may be 20% of the received data.
The structure of the optimum pest forecasting model structure and its parameters for each field may get modified based on error between actual observation and the predicted outcome. The system 100 of the present disclosure adapts the generated pest forecasting model with current scenario of pests and improves the model with time. In one embodiment, the current scenario of pest infestation are based on on-field observation captured using samples collected from farms through participatory sensing platform. If the pest forecasting model outcome does not match with the actual pest infestation in farms, then the error is adapted in the model. The on-field observations may include sensor data from wireless sensor networks and events captured by mobile phone applications (including farm operations, pesticide spray activities or any other information that can uniquely identify a field). Wireless Sensor Networks (WSNs) have a base station and different sensor nodes with sensors capturing hourly temperature, humidity, soil moisture, leaf wetness, light intensity, etc. Sensed parameters from different sensor nodes can be aggregated for the geo-location under consideration. Thus, in an embodiment, the one or more processors 104 further validate and adapt the optimum pest forecasting model based on reported first set of inputs. For instance, the reported first set of inputs may comprise one or more of participatory sensing inputs and crowd sourcing inputs pertaining to the agronomic information.
In an embodiment, the one or more processors 104 may validate and adapt the optimum pest forecasting model by first receiving images from the reported first set of inputs. The received images are filtered to extract region of interest (ROI) in the images. In an embodiment, the ROI may be a portion of vegetation like leaves or stem that has pest/disease patches therein in the images, the disease patches being pest infected patches. The ROI may be then analyzed for one or more features such as size, color, shape, texture, and the like. Based on the analyzed features, maturity of pest/disease may be computed and the optimum forecasting model can be adapted accordingly. In an exemplary embodiment, the reported first set of inputs may include images of infected leaves/stem and time stamp of image capture from fields through a Rural Participatory Sensing Application. Rural participatory sensing makes large scale validation and adaption of developed forecasting model. The received images may be filtered, pest/disease patches on the infected leaves/stem may be identified and image analysis may be performed to compute maturity of pest at the time of image capture. From the information pertaining to maturity of pest/disease, actual pest infestation instances may be derived. For instance, if pest is of age 10 days, actual pest infestation instance is taken as 10 days back from the date of event reporting.
In an embodiment, pseudo code for image analysis may be represented as—
In an embodiment, pseudo code for validating and adapting the optimum pest forecasting model may be represented as—
In an embodiment, the reported first set of inputs may be stored in a database for further processing. The database may also contain events reported from participatory sensing platforms. Farmers may use a mobile phone application for capturing on-field events. Exemplary event attributes that may be captured include geo-tagged images of pest or disease affected vegetation, pest counts on each plant or leaf, number of plants damaged in farms, number of leaves damaged, severity of pest or disease, percentage of leaf area damaged, and the like. Using this information, the system of the present disclosure can suitably update the pest forecasting model.
Following are the cases when the performance of the system of the present disclosure may get affected:
In an embodiment, the system of the present disclosure may propose incentive mechanisms for its users. Incentives may be decided based on the performance of users with regards to timely response to disease/pest infestation, response to requests for periodic sharing of information, and the like.
In an embodiment, the system and method of the present disclosure provides a digitized farm diary which may keep record of all farming activities for future references for a specific scenario. Pest or disease incidences and corresponding chemical usages in each farm or site can be tracked and recorded. This will help farmers to decide on the control measures to be taken in future. Likewise, different farming activities like irrigation, fertilizer application, control measures taken for a particular pest of disease infestation scenario can also be tracked which will help farmers to generate similar yield or improve the yield in future instances.
The written description describes the subject matter herein to enable any person skilled in the art to make and use the embodiments of the present disclosure. The scope of the subject matter embodiments defined here may include other modifications that occur to those skilled in the art. Such other modifications are intended to be within the scope if they have similar elements that do not differ from the literal language of the claims or if they include equivalent elements with insubstantial differences from the literal language.
It is, however to be understood that the scope of the protection is extended to such a program and in addition to a computer-readable means having a message therein; such computer-readable storage means contain program-code means for implementation of one or more steps of the method, when the program runs on a server or mobile device or any suitable programmable device. The hardware device can be any kind of device which can be programmed including e.g. any kind of computer like a server or a personal computer, or the like, or any combination thereof. The device may also include means which could be e.g. hardware means like e.g. an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or a combination of hardware and software means, e.g. an ASIC and an FPGA, or at least one microprocessor and at least one memory with software modules located therein. Thus, the means can include both hardware means and software means. The method embodiments described herein could be implemented in hardware and software. The device may also include software means. Alternatively, the embodiments of the present disclosure may be implemented on different hardware devices, e.g. using a plurality of CPUs.
The embodiments herein can comprise hardware and software elements. The embodiments that are implemented in software include but are not limited to, firmware, resident software, microcode, etc. The functions performed by various modules comprising the system of the present disclosure and described herein may be implemented in other modules or combinations of other modules. For the purposes of this description, a computer-usable or computer readable medium can be any apparatus that can comprise, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device. The various modules described herein may be implemented as software and/or hardware modules and may be stored in any type of non-transitory computer readable medium or other storage device. Some non-limiting examples of non-transitory computer-readable media include CDs, DVDs, BLU-RAY, flash memory, and hard disk drives.
Further, although process steps, method steps, techniques or the like may be described in a sequential order, such processes, methods and techniques may be configured to work in alternate orders. In other words, any sequence or order of steps that may be described does not necessarily indicate a requirement that the steps be performed in that order. The steps of processes described herein may be performed in any order practical. Further, some steps may be performed simultaneously.
The illustrated steps are set out to explain the exemplary embodiments shown, and it should be anticipated that ongoing technological development will change the manner in which particular functions are performed. These examples are presented herein for purposes of illustration, and not limitation. Further, the boundaries of the functional building blocks have been arbitrarily defined herein for the convenience of the description. Alternative boundaries can be defined so long as the specified functions and relationships thereof are appropriately performed. Alternatives (including equivalents, extensions, variations, deviations, etc., of those described herein) will be apparent to persons skilled in the relevant art(s) based on the teachings contained herein. Such alternatives fall within the scope and spirit of the disclosed embodiments. Also, the words “comprising,” “having,” “containing,” and “including,” and other similar forms are intended to be equivalent in meaning and be open ended in that an item or items following any one of these words is not meant to be an exhaustive listing of such item or items, or meant to be limited to only the listed item or items. It must also be noted that as used herein and in the appended claims, the singular forms “a,” “an,” and “the” include plural references unless the context clearly dictates otherwise.
It is intended that the disclosure and examples be considered as exemplary only, with a true scope and spirit of disclosed embodiments being indicated by the following claims.
| Number | Date | Country | Kind |
|---|---|---|---|
| 201621023796 | Jul 2016 | IN | national |
