INGENIEUR
INGENIEUR
performed by manually counting the total number of rice plant tillers in the square frame . This sampling technique is time-consuming , labour intensive and costly [ 13 ]. To solve this problem , Teoh et al . ( 2008 ) [ 14 ] introduced an automatic counting plant tiller method using an image processing technique . Digital images of rice plants in the square frame were captured and classified into plant and non-plant regions using an image classification method . Linear regression analysis results indicated that a good correlation coefficient value of 0.8328 existed between the plant tiller count and the area of the plant region . Verification results also showed that the linear model is capable of estimating tiller numbers with 92.17 % average accuracy . However , both methods are only applicable for tiller counting on the ground in small fields [ 13,14 ].
To solve the problem , a UAV equipped with a camera has been used for counting rice plants [ 15 ]. The results revealed that the correlation of the rice plant count between the remote sensing method and ground counting method is R 2 = 0.893 . Hence , UAV-based remote sensing is capable of tallying plant numbers for generating a plant population density map .
FERTILISER MANAGEMENT
Variable-rate nitrogen ( N ) management is one of the most important activities in PA for minimising fertiliser input and maximising yield . Traditional N application methods such as blanket or package fertiliser recommendations over large areas are not efficient because indigenous nutrient supply varies widely in rice fields [ 16 , 17 ]. According to Peng et . al . ( 1996 ) [ 18 ] and Balasubramanian et . al . ( 1999 ) [ 19 ], a SPAD chlorophyll meter can be used to monitor plant N status in situ in the field and determine the right time for N topdressing in rice . However , field manual measurement of plant N content using the SPAD chlorophyll meter is time-consuming and tedious .
To overcome the problem , low-altitude remote sensing using a camera attached to a UAV was applied to improve the manual measurement of plant N status in a larger paddy field . For example , Vegetation Indices ( VIs ) have been used in remote sensing applications for monitoring the crop N status in paddy fields [ 20 - 23 ]. The dark green colour index ( DGCI ) derived from aerial images was used to assess the N concentrations in the field [ 20 ]. It was found that DGCI values predicted the N concentrations with R 2 = 0.672 . The band ratio Red /( Red + Green + Blue ) and Normalised Difference Vegetation Index ( NDVI ) were used to estimate SPAD values [ 21 , 22 ]. The results demonstrated that the Red /( Red + Green + Blue ) and NDVI produced highly accurate SPAD estimations with r 2 = -0.97 and R 2 = 0.76 , respectively . According to Zheng et al . ( 2018 ) [ 23 ], the Red Edge Vegetation Indices ( REVI ) calculated from multispectral images produced high estimation accuracy for leaf nitrogen accumulation ( range of R 2 : 0.79 - 0.81 ) and plant nitrogen accumulation ( range of R 2 : 0.81 - 0.84 ). The results documented in previous reports indicated that the UAV-based N status monitoring system is capable of providing useful information for near real-time decisionmaking on variable-rate N management in paddy fields [ 20-23 ].
PEST , DISEASE AND WEED MANAGEMENT
Pest , disease and weed ( P & D & W ) are very important factors that can cause significant economic loss due to reduced yield and quality . Crops should be monitored constantly to detect the P & D & W in time and avoid spreading problems . Traditionally , monitoring and blanket spraying of chemical tasks are performed manually in the field . However , this can be very time-consuming , labour intensive and costly .
To overcome the problems , precision agriculture adopts a variable rate technology ( VRT ) that has been used for controlling the P & D & W in the paddy field . VRT refers to the spatially variable application of chemicals according to site-specific requirements . P & D & W distribution maps are important in VRT for precise spraying of chemicals . Applications of UAV-based remote sensing technologies for P & D & W monitoring and mapping in the paddy fields have been reported by Yue et al . ( 2012 ) [ 24 ], Zhang et al . ( 2018 ) [ 25 ], Huang et al . ( 2018 ) [ 26 ] and Stropping et al . ( 2018 ) [ 27 ], respectively . A distributional map of crop pest status has been generated and classified into healthy ,
52 VOL 99 JULY - SEPTEMBER 2024