Sorry, you need to enable JavaScript to visit this website.
Skip to main content

PalmWatch: monitoring palm trees using remote sensing

PalmWatch: Monitoring of Palm Tree Farms with Advanced Remote Sensing Techniques

 

Project number: 10-SPA1192-02 
Project Investigator: Dr. Naif Alajlan
Period: Jan 2012- Jun 2014

 

Team members:

Project Summary:

Palm trees cultivation continues to be the major agricultural cop of Saudi Arabia. The government considers palm trees as a national symbol and dates as critical for national food security. The availability of precise information about palm tree farms in an appropriate time interval is very relevant for a large number of socio-economic decisions. 

This research project, proposes to develop a set of remote sensing technologies for monitoring palm tree farms in Saudi Arabia. In particular, it aims at proposing innovative methods to address four issues of major interest to palm farmers. These are: 1) palm tree counting; 2) palm crop forecasting; 3) change detection; and 4) multitemporal classification for updating land-cover maps. 

To support the goals of the NPST program in the strategic technology are of space and aeronautics, this project aims at reaching the following targets: 
1) Contributing in the increase of exploitation of remote sensing data in agricultural and crop management applications of potential interest to local private and public agricultural institutions; and 
2) Contributing in the evolvement of remote sensing research area in Saudi Arabia. 

This project is important for the field of palm tree cultivation in Saudi Arabia and other countries, because:

 

1.      It Offers great opportunities to the modernization of the procedures of farming palm trees in Saudi Arabia. The findings could be extended to monitor other important agricultural products, such as mango, olive trees, and others.

2.      It improves the effectiveness of the incentive programs provided by the government to support and manage the huge number of palm trees in the kingdom (approximately 23 million).

3.      It improves the quality of dates and date products.

4.      It promotes the remote sensing area at both academic and research levels, and contribute to its beneficial use in the economy of the country.

5.      It consolidate Saudi academic and research know-how on the rapidly evolving area of remote sensing and precision farming.

 

Project objectives:

1. Palm tree counting: The information about the number of palm trees is essential for surveying and inventorying and relevant for predicting the crop production. In addition, the number of trees is used as a fundamental criterion for obtaining public grants. Usually, the process of counting palm trees is carried out manually on site. Such an operation is very tedious due to the typically high number of plantations to monitor. Therefore the utilization of automatic methods, which exploit images acquired by very-high-resolution sensors, represents for the public administration a potentially interesting alternative. 

2. Palm crop forecasting: The problem of crop production forecasting is important for agricultural management, food security warning at regional and national levels. On a field scale, it plays an important role in determining the farm income and household food consumption. Crop estimates based on field reports are often expensive, prone to large errors, and cannot be real time spatially explicit estimates. The availability of spatial and temporal information from RS sensors makes it a very attractive tool to efficiently carry such tasks by studying the effect of the environmental conditions on palm crop production. 

3. Unsupervised change detection: Detecting changes occurred between two or more acquisition dates over the same geographical area represents one of the most interesting applications of RS. Change detection techniques are very useful in multitemporal monitoring of palm tree farms status by discerning areas of changes due to environment conditions or human activities. Such information could be obtained without the necessity of collecting any ground-truth data for all acquisition dates making it an efficient data-analysis tool. Usually, the output of change detection methods is a basic thematic map characterized by change and no-change classes. 

4. Multitemporal classification for updating land-cover maps: When more precise information is needed such as the generation of accurate multitemporal classification maps for land-cover updating instead of basic-change detection maps, it becomes necessary to integrate prior knowledge (ground-truth information) about the study areas to be monitored. The availability of such prior knowledge (for one or more acquisition dates) allows the development of sophisticated approaches based on supervised or semisupervised classification to produce accurate classification maps. 

 

Supporting NSTIP goals:

 

The current research project supports the NSTIP main goals and objectives in the following 4 ways:

 

STRATEGIC GOALS OF NSTIP TECHNOLOGY PROGRAM

PROJECT SUPPORT TO GOALS

To develop into the leading service provider of commercial Earth Observation products within the region.

One of the main outcomes of the project are Remote sensing algorithms and modules (change detection, palm tree counting, multitemporal classification)

The availability of these product will highly contribute towards developing a leading service provider of commercial Earth Observation products within the region.

To raise the level of aerospace higher education and training programs within the Kingdom of Saudi Arabia (KSA) and to expand interest and resources in the space and aeronautical sectors.

During the project two undergrad graduation projects in the RS technology program area are supervised (two students in each project) as well as one masters student.

To promote the wider national use of space and aeronautical projects and services within government, industry, and the general public.

This project promotes the ministry of agriculture and other public and private entities to use of RS technology in precision farming.

A workshop is organized to disseminate the results into for the public and private sectors.

To become a participant in international or regional space and aeronautic science missions.

Through publications in reputable international journals and conferences, the role of Saudi Arabia within the international RS community is promoted.

 

Project progress:
 

 

 

Objectives

Phases

Tasks

Status (Completed, Ongoing, Planned, Modified *, Discontinued*)

Percentage of achievement

O1: Tree counting and

O2: Palm crop forecasting and

O3: Change detection

O4: Multitemporal classification for updating land-cover maps

1

T 1.1 Site identification and RS data purchasing

Completed

100%

T 1.2 Data preparation

Completed

100%

T 1.3 Update of the literature

Completed

100%

T 1.4. System structure of PalmWatch

Completed

100%

T 1.5 Method and algorithm Design

Completed

100%

T 1.6 Technical and coordination meeting 

Completed

100%

T 1.7 Progress report writing

Completed

100%

O1: Tree counting and

O3: Change detection

2

T 2.1 Software implementation

Completed

100%

T 2.2 Experimental validation

Completed

100%

T 2.3 Results dissemination

Completed

100%

T 2.4 Technical and coordinating meeting

Completed

100%

T 2.5 Progress report writing

Completed

100%

O2: Palm crop forecasting and

O4: multitemporal classification for updating land-cover maps

3

T 3.1 Software implementation

Completed

100%

T 3.2 Experimental validation

Completed

100%

T 3.3 Technical and coordinating meeting

Completed

100%

T 3.4 Progress report writing

Completed

100%

O2: Palm crop forecasting and

O4: multitemporal classification for updating land-cover maps

4

T 4.1 Results dissemination

Completed

100% 

T 4.2 Software and system architecture refinement

Completed

100% 

T 4.3 Technical and coordinating meeting

Completed

 100%

T 4.4 Final report writing

Completed

 100%

 

 

 

 

 

 

 

Where the phases are as outlined in the original proposal are as follows:

 

PHASE I: Methodological development and RS data purchasing

Task 1.1 Methodological development

Task 1.2 RS data purchasing

Task 1.3 Progress report writing

PHASE 2: Implementation (tree counting and change detection)

Task 2.1 Software implementation

Task 2.2 Experimental validation

Task 2.3 Results dissemination

Task 2.4 Technical and coordinating meeting

Task 2.5 Progress report writing

PHASE 3: Implementation (forecasting and multitemporal classification)

Task 3.1: Software implementation

Task 3.2: Experimental validation

Task 3.3: Technical and coordinating meeting

Task 3.4: Progress report writing

PHASE 4: RS system refinement

Task 4.1 Results dissemination

Task 4.2 Software and system architecture refinement

Task 4.3 Technical and coordinating meeting

Task 4.4 Final report writing

 

Project contributions to knowledge:

 

The contributions of the project can summarized as follows:

 

1.     The project proved the feasibility of a system for palm tree counting and monitoring using satellite images or UAV images.

2.     Developed successfully new system for the remote monitoring of palm tree farms, which includes :

·       Palm tree counting module which provides a web based GUI, which enables interested public and private parties to use this service. Through the web portal they can get more information about the research findings of this project and request the service of palm tree counting from us. The website can be accessed at http://www.alisr.com/projects/palm_watch

·       Palm forecasting module: to predict the date palm production in a palm field

·       Change detection module to detect changes in palm field between two acquisition dates  due to incorrect farming or due to switching to another crop.

·       Multitemporal classification module for classifying images acquired at different times and different fields into palm and no-palm classes.

 

3.     Published three (3) papers in top ISI remote sensing journals, and five (5) papers in international conferences.

4.     Developed a method for general estimation problems based on Gaussian process regression. This method is also suitable for date palm crop estimation. The results has been published in IEEE Transactions on Geoscience and remote sensing, 2012 (IF: 3.47)   

5.     Developed a method for interactive change detection in VHR images (published in IEEE Geoscience and remote sensing letters, 2014 (IF: 1.82)).

6.     Developed two methods for counting palm trees in VHR and UAV images, respectively. The second method related to UAV has been accepted recently for publication in the IEEE journal  of selected topics in applied earth observation and remote sensing (IF: 2.87). These two methods are also suitable for the Multitemporal classification module.

7.     Published a summary about the project at the college of computer and information sciences newsletter (which was distributed by hardcopy and by email) to inform students and faculty at the college about the project. This has helped attract four (4) undergrad students who ended up performing two (2) undergraduate projects in the topic of the research project besides an M.Sc student.

8.     We have invited Dr. Farid Melgani, the consultant, to give a seminar at King Saud University about the latest advanced in remote sensing and its applications. This helped attract several graduate students who are interested to do their graduate studies in remote sensing.

9.     Organized a workshop for result dissemination, and getting feedback from the community.

 

Project outcomes:

Algorithms and software packages:

1) RS Palm tree farm database collection
2) Estimation algorithm for crop forecasting
3) Change detection module
4) Palm tree counting Web portal  


Journal publications:
 

1.      Malek, Y. Bazi, N. Alajlan, H. Hichri, and F. Melgani, “Efficient framework for palm tree detection in UAV images”, IEEE Journal selected topics in applied earth observation and remote sensing (JSTARS) (IF: 2.87).  (accepted), 2014.    

            Click here to view paper in PDF

 

2.      Haikel AlHichri, Yakoub Bazi, Naif Alajlan, and Salim Malek, "Interactive segmentation for change detection in multispectral remote-sensing images," IEEE Geoscience Remote Sensing Letters, vol. 10, no. 2, pp. 298-302, March 2013. (IF: 1.56).  

Link:  http://ieeexplore.ieee.org/xpl/articleDetails.jsp?tp=&arnumber=6244846&contentType=Journals+%26+Magazines&sortType%3Ddesc_p_Publication_Year%26searchField%3DSearch_All%26queryText%3DAlajlan

 

3.      Yakoub Bazi, Naif Alajlan, and Farid Melgani, "Improved estimation of water chlorophyll concentration with semisupervised Gaussian process regression," IEEE Transactions on Geoscience and Remote Sensing, vol. 50, no. 7, pp. 2733-2743, July 2012. (IF: 3.47)

Link:

http://ieeexplore.ieee.org/xpl/articleDetails.jsp?tp=&arnumber=6122062&contentType=Journals+%26+Magazines&sortType%3Ddesc_p_Publication_Year%26searchField%3DSearch_All%26queryText%3DAlajlan

        4.     H. Hichri,  Y. Bazi,  N. Alajlan,  N. Ammour,  and S. Malek, , Clustering of Hypersectral images with an ensemble method based on fuzzy c-means and Markov random fields, Arab Journal of Science and Engineering, vol. 39, no. 9, P 3747-3757, May 2014 (ISSN 1319-8025). 

http://apps.webofknowledge.com/full_record.do?product=UA&search_mode=GeneralSearch&qid=2&SID=Q2d98m7TzLmalBAqLfW&page=1&doc=2

 

Conference Publications:

5.      H. Hichri “SIFT-ELM Approach for unsupervised change detection in VHR images," IEEE- IGARSS-2014, Quebec city, Canada, July 2014.

 

http://www.igarss2014.org/Papers/PublicSessionIndex3.asp?Sessionid=1267

 

6.      S. Malek, Y. Bazi, N. Ajlan, H. Hichri, “An automatic Approach for palm tree counting in UAV images," IEEE- IGARSS-2014, Quebec city, Canada, , July 2014.

 

               http://www.igarss2014.org/Papers/viewpapers.asp?papernum=3813

 

7.      Alajlan, Naif, "Classification of VHR Images Based on SVM and Multiobjective Evolutionary Optimization," Computational Science and Computational Intelligence (CSCI), 2014 International Conference on , vol.1, no., pp.194,198, 10-13 March 2014

8.      Naif Alajlan, Yakoub Bazi, Haikel Hichri, and Essam Othman, "Robust classification of hyperspectral images based on the combination of supervised and unsupervised learning paradigm," Proc. of the IEEE International Geoscience and Remote Sensing Symposium IGARSS-2012, pp. 1417-1420, Munich, Germany, July 2012.

Link:

http://ieeexplore.ieee.org/xpl/articleDetails.jsp?tp=&arnumber=6351270&contentType=Conference+Publications&sortType%3Ddesc_p_Publication_Year%26searchField%3DSearch_All%26queryText%3DAlajlan

 

9.      Haikel Hichri, Yakoub Bazi, Naif Alajlan, and Sayed M. Ahmad, "Interactive change detection techniques in multitemporal multispectral remote-sensing images," Proc. of the IEEE International Geoscience and Remote Sensing Symposium IGASS-2012, pp. 6173-6176, Munich, Germany, July 2012.

Link:

http://ieeexplore.ieee.org/xpl/articleDetails.jsp?tp=&arnumber=6352666&contentType=Conference+Publications&sortType%3Ddesc_p_Publication_Year%26searchField%3DSearch_All%26queryText%3DAlajlan

 

 Awards:

 

1) Silver medal in

The Malaysia tech Invention Expo (www.mte.com.my)

Mar 2012

Contact Us

For information about the project, contact:

The mailing address for ALISR is:

College of Computer and Information Sciences
PO Box 51178
King Saud University 
Riyadh, Saudi Arabia  11543

Last updated on : January 12, 2023 1:11am