Agricultural Research Management Information System

Home
Research Summary
Search
About Us
- ARMIS
- Brochure
Contact Us
Report
User Request
Data Input
Help
- Operation Manual
  - PDF
  - Video
- Program Area & Commodity
We have reached 37600 number of research entries at this moment.

- Logout

Research Detail

Home
Research
Detail

Image-based Plant Disease Detection: an Experience of Bangladesh

Md. Abdul Malek
Department of Agricultural Extension and Information System, Sher-e-Bangla Agricultural University, Sher-e-Bangla Nagar, Dhaka-1207

Md. Sekender Ali
Department of Agricultural Extension and Information System, Sher-e-Bangla Agricultural University, Sher-e-Bangla Nagar, Dhaka-1207

Mst. Fauzia Sultana
Department of Parasitology, Bangladesh Agricultural University, Mymensingh

Abstract/ Executive Summary:

Use of image-based plant disease detection system (IPDDS) is getting popularity day by day. In this case images of plant diseases are used to detect the symptom. Several sophisticated techniques are used in IPDDS. But in the recent year Bangladesh has experienced a very simple solution called krishoker janala (farmer's window) in providing agricultural advisory services to the farmer. It is actually an imagebased system (mobile application also available) of plant disease detection. Farmer/ user can easily real-time identify problem of their plant by matching the symptom of the plant with the image of the system. A case study was conducted to find out the effectiveness of krishoker janala as perceived by the user. The study reveals that 84.62%, 11.54% and 3.85% respondent found kriskoker janala as highly effective, moderately effective and less effective respectively. The study also found that 92% respondents could solve their plants' problem using krishoker janala application; did not need help from others. That was an exceptional experience of Bangladesh in the field of image-based plant disease detection. Study also reveals that user reported some problems regarding the mobile application and they also provided some recommendations to increase the effectiveness of the system. These recommendations to be considered for increasing the effectiveness of this system during next updating. Policy maker of the country should include IT-industry into collaboration to incorporate automate system.

Keywords: Krishoker janala, Farmer's window, Image-based system, Plant disease identification

Location: In Bangladesh

Start Date:

End Date:

Program Area: Pest Management

Commodity: Diseases

Objectives:

The main objective of the study was to assess the effectiveness of the krishoker janala app for providing agricultural services to the farmers. Specific objectives of the study were as follows:

To explore farmer's understanding on krishoker janala mobile application
To find out the benefits that farmers are getting through using the mobile application
To identify the challenges faced by users while using the app
To analyze krishoker janala as a service
To measure the satisfaction level of the user farmer
To find out improvement scope of the mobile application desired by the user.

Methodology:

The study was conducted as a case study. Although case study research method is often considered to be invalid, invaluable and improper (Yin, 2003), it is an established research method in social sciences (Teegavarapu et al., 2008) -a review of literature in social sciences (Yin, 2003; Lukoff et al.,1998 and Dresch et al., 2015) support the argument. The study was conducted in two Upazilas viz. Sadar upazila of Norsindi and Sadar upazila of Gazipur. The Upazilas were selected based on the user density report provided by the Department of Agriculture Extension (DAE). The top two user density Upazilas were selected for the study. Two Focus Group Discussions (FGDs) were conducted (10 participants each). Key Informant Interview was conducted with Upazila Agriculture Officer, Agriculture Extension Officer and Sub-Assistant Agriculture Officer to get primary data. Secondary documents, media reports etc. were thoroughly studied. The content of the website, application and user responses were also studied. Based on FGDs, Key Informant Interview and review of literature and secondary documents, interview schedule was prepared. Users of almost all categories were included in the study. Out of 3328 registered users a total of 330 users were selected randomly from krishoker janala app user registration information and interviewed. Reviews of the app user were also collected from Google play store. After completing the collection of data, data were cleaned and analyzed and graphs and charts were presented in excel.

Some limitations of the study included: firstly- interviewee from krishoker janala app users registration information, were those users who had registered themselves for sending picture to the application. (krishoker janala has got a picture sending option). Secondlyreviewers of the app are advanced level users. Ensuring proportionate sampling was a challenge of the study. To overcome the limitations and ensuring Guba's (Guba, 1981) criteria (credibility, transferability, dependability and confirmability) for the study; various techniques were applied such as consultation of appropriate documents and preliminary visits to the respondents and authority (Lincoln et al.,1981) were done and good relationship (Erlandson et al.,1993) was established between researcher and users. Triangulation (Maanen, 1983) was done by the use of a wide range of informants. The use of “overlapping methods” as suggested by Shenton et al. (2004) was done by conducting the FGDs and personal interview. Despite all these steps, the presence of advanced user was dominating to some extent in this study. Therefore, scopes prevail for further study to know the opinion of common users of kriskoker janala mobile application. Regarding using case study as research methods it can be concluded with the help of Yin, the most prominent critic of this method that case study is not only a suitable method to explore the subject; it also can be used to explain certain phenomena (Yin, 1984).

Source of Information: Bangladesh Journal of Extension Education, Volume 32: 2020 (Special Issue)

Article Link (Online Journal):

Funding Source:

Budget:

Total Budget (Tk.) :

Achievement/Findings:

The study reveals that 84.62%, 11.54% and 3.85% user respondents found kriskoker janala application as highly effective, moderately effective and less effective, respectively. The application found to be effective in detecting plant's problem and it also saved time, cost and number of visit of farmers in getting agricultural services. The study also found that 92% respondents could solve their plants' problem using krishoker janala application; did not need help from others. These figures give a clear indication that krishoker janala application had created capacity among the users to solve the problems of their crops. It was also seen that almost all the respondents (94%) got timely service by using krishoker janala application. It was an exceptional experience of Bangladesh in the field of image-based plant disease detection. The study also found that user respondents reported some problems like content not updated, the system is not automated, difficulties in using with updated android version (android 8.0 and above), faulty registration system, bigger size of the app etc. they also provided some recommendations for further improvement of the application like campaigning and promotion of the application, making available for all platform (e.g. iOS, windows) to cover more user group, incorporating feedback system and sharing option so that user can provide feedback and share the content to the social media, updating the information regularly, providing information on production technology automation of disease detection system, etc. Recommendations should be taken to action by the policymaker of the country to make the application more effective to the users. To incorporate automatic disease detection systems IT industry should be included in the collaboration.

Knowledge Management: Journal