Search

The interest in visual-based surveillance systems, especially in natural disaster applications, such as flood detection and monitoring, has increased due to the blooming of surveillance technology. In this work, semantic segmentation based on convolutional neural networks (CNN) was proposed to identify water regions from the surveillance images. This work presented two well-established deep learning algorithms, DeepLabv3+ and SegNet networks, and evaluated their performances using several evaluation metrics. Overall, both networks attained high accuracy when compared to the measurement data but the DeepLabv3+ network performed better than the SegNet network, achieving over 90% for overall accuracy and IoU metrics, and around 80% for boundary F1 score (BF score), respectively. When predicting new images using both trained networks, the results show that both networks successfully distinguished water regions from the background but the outputs from DeepLabv3+ were more accurate than the results from the SegNet network. Therefore, the DeepLabv3+ network was used for practical application using a set of images captured at five consecutive days in the study area. The segmentation result and water level markers extracted from light detection and ranging (LiDAR) data were overlaid to estimate river water levels and observe the water fluctuation. River water levels were predicted based on the elevation from the predefined markers. The proposed water level framework was evaluated according to Spearman’s rank-order correlation coefficient. The correlation coefficient was 0.91, which indicates a strong relationship between the estimated water level and observed water level. Based on these findings, it can be concluded that the proposed approach has high potential as an alternative monitoring system that offers water region information and water level estimation for flood management and related activities.

Flood disasters are considered annual disasters in Malaysia due to their consistent occurrence. They are among the most dangerous disasters in the country. Lack of data during flood events is the main constraint to improving flood monitoring systems. With the rapid development of information technology, flood monitoring systems using a computer vision approach have gained attention over the last decade. Computer vision requires an image segmentation technique to understand the content of the image and to facilitate analysis. Various segmentation algorithms have been developed to improve results. This paper presents a comparative study of image segmentation techniques used in extracting water information from digital images. The segmentation methods were evaluated visually and statistically. To evaluate the segmentation methods statistically, the dice similarity coefficient and the Jaccard index were calculated to measure the similarity between the segmentation results and the ground truth images. Based on the experimental results, the hybrid technique obtained the highest values among the three methods, yielding an average of 97.70% for the dice score and 95.51% for the Jaccard index. Therefore, we concluded that the hybrid technique is a promising segmentation method compared to the others in extracting water features from digital images.

Flooding is an annual event in Malaysia that causes damage of properties and loss of life. There are still lack of information available to explain the impact of flooding in the agricultural sector. Thus, this paper focuses on flooding and its impact on oil palm plantations. A flood model is developed by using InfoWorks ICM to determine the impact of floods before and after mitigation measure. Flood mapping and Geographical Information System (GIS) technology are integrated to determine the flood losses on agricultural land along the Selangor River, specifically in oil palm plantations. The results show that the estimated total damage losses for two flood events; 10 year average recurrence intervals (ARI) and 100 year ARI, involved millions of ringgits. A retention pond is proposed, modeled and tested as a flood mitigation option. The mitigation measure reduced the flood inundation in the agricultural area significantly by 64.89% for 10 year ARI and 72.09% for 100 year ARI. Besides, the losses in oil palm plantation are also reduced significantly by 64.52% for 10 year ARI and 72% for 100 year ARI.

Nowadays, smart phone has become a necessity as it offers more than just making a phone call. Smart phone combines the features of cell phone with other mobile devices such as personal digital assistant (PDA) and GPS navigation unit that propel the popularity of smart phones. In recent years, the interest in mobile communication has been increased. Previous research using mobile application has been successfully done in varies areas of study. Areas of study that have been done are health care, education, and traffic monitoring. Besides, mobile application has also been applied in agricultural sector for various purposes such as plant pest risk management. In this study, mobile application for data collection on Ganoderma disease of oil palm has been successfully developed. The application uses several devices in a smart phone such as GPS, Wifi/ GPRS connection and accelerometer devices. The application can be installed in the smart phone and users can use the application while working on-site. The data can be updated immediately through their smart phones to the service. Besides, the application provides offline map so the user can be productive even though their network connectivity is poor or nonexistent. The data can be synced when the users online again. This paper presents an application that allows users to download features from a sync-enabled ArcGIS Feature Service, view and edit the features even when the devices fail to connect with any network connectivity while collecting data on-site.

Flooding is the most significant natural hazard in Malaysia in terms of population affected, frequency, flood extent, flood duration and social economic damage. Flooding causes loss of lives, injuries, property damage and leave some economic damage to the country especially when it occurs in a rural area where the main income is dependent on agricultural area. This study focused on flooding in oil palm plantations, rubber plantations and fruits and vegetables area. InfoWorks ICM was used to develop a flood model to study the impact of flooding and to mitigate the floods using a retention pond. Later, Geographical Information System (GIS) together with the flood model were used for the analysis on flood damage assessment and management of flood risk. The estimated total damage for three different flood event; 10 ARI, 50 ARI and 100 ARI involved millions of ringgits. In reducing the flood impact along the Selangor River, retention pond was suggested, modeled and tested. By constructing retention pond, flood extents in agricultural area were reduced significantly by 60.49% for 10 ARI, 45.39% for 50 ARI and 46.54% for 100 ARI.

Plant diseases contribute to significant economic and post-harvest losses in agricultural production sector all over the world. Early detection of plant diseases and pathogens is crucial for maintaining the sustainability of the agricultural sector. In this regard, non-destructive techniques are one of the most practical and feasible ways of monitoring health conditions of plants for real-time applications. Current methods that involve molecular and serological methods are time-consuming, laborious, and require complex procedures. Hence, non-destructive methods are preferable since no sample preparation and repetitive processes are needed for the measurements. The prevention of plant disease and pathogens during the preliminary stage aids to the health control and yield optimisation without depending on the pest and pesticides usage. These techniques are boosted by the development of advanced technologies where computer-aided systems are employed for monitoring plant diseases. It is also vital to analyse the disease spread in plants for overcoming the issues related to physiological and biological states in crop protection. Most often a wide range of non-destructive techniques is needed to provide rapid detection of plant infections even in the case of one plant type. This study reviews the applications of non-destructive techniques including image processing, imaging-based, spectroscopy-based and remote sensing methods in monitoring and detecting plant diseases. The non-invasive techniques have also been investigated for their fundamental concepts and potential applications in developing sensing devices for fieldwork.

Flooding is the major natural hazard that occurs in Malaysia. Flooding causes loss of lives, injuries, property damage and leave economic damage to the country, especially when it occurs in an agricultural area. There is still a lack of information available on floodplain especially on the impact of flooding in the agricultural area. This study focused on flooding and flooding impact in oil palm plantations, fruits and vegetables area. A river modeling is required to study the impact of flooding and to mitigate the floods using one mitigation option. A flood model was developed using InfoWorks Integrated Catchment Model (ICM) to carry out the analysis for flood damage assessment. With the aim of creating a flood damage map, Geographical Information System (GIS) was combined with the flood model to provide an ideal tool for the analysis of the flood damage and the effect of mitigation to flood damage. The estimated total damage for three different flood event; 10 ARI, 50 ARI and 100 ARI involved millions of ringgits. In order to reduce the flood impact along the Selangor River, a flood mitigation structure which is a retention pond was suggested, modeled and simulated. The effects of the retention ponds were analyzed and evaluated for 10 ARI, 50 ARI and 100 ARI. With this retention pond, flood extents of the flood events in agricultural area were shown capable of reduction significantly by 65.57% for 10 ARI, 76.18% for 50 ARI and 72.51% for 100 ARI.