Maximizing a Fruit Plantations Yield using UAV Imagery
Plant Counting and Gap Filling Techniques
Determining plant density and identifying canopy gaps is crucial for good plantation management; this helps predict yields and to maximize the yield by refilling planting gaps. Scientific photographs taken from Unmanned Aerial Vehicles (UAVs) allow accurate plant counts and the identification of plant mortality for several hundreds of hectares per day. Studies over different crops show that QuestUAV images are a powerful tool for a plantation management to develop replanting plans and to compare actual vs. target plant density.
Healthy Plantation vs High Mortality
Numerous studies have been conducted on the effects of plant density on growth and yield of tropical crops. Plant densities are an important and limiting factor for growth, nutritional status, fruiting and hence for a plantation’s yield. Optimal plant densities depend on different factors, such as cultivars, climate, soil characteristics, land preparation etc.
Low-quality planting material, wrong plant treatment or climatic anomalies can lead to high mortality. The result can be a huge reduction of yield rates. Further, actual plant densities can differ from target densities for several reasons, such as wrong distances between planting rows or a lack of planting material. Therefore, the refilling of canopy gaps and correction of non-optimal plant densities are of high priority for a good plantation management.
Case Study: Developing Management Plans on the Basis of UAV Images
Several flights were conducted with the QuestUAV Surveyor-Pro over different tropical crops in Thailand, Indonesia and the Philippines. More than 1,000 ha of pineapple, banana and oil palm were mapped by our crew. All images were processed with Pix4Dmapper Pro and analyzed by plantation management with the open-source software QGIS. The software allows plant counting, density calculations and the generation of mortality maps by visual inspection of the image products. More advanced approaches, like image classification and pattern detection algorithms, allow to map canopy gaps and determine plant densities in an automatic or semi-automatic way.
Identifying and Refilling Canopy Gaps
The figure below shows a study with the objective to map canopy gaps and mortality at a banana plantation in Indonesia. The analysis was done by the plantation management in QGIS. Yellow circles represent missing or dead banana plants, digitally marked by visual inspection of the image products. The map is the basis for the plantation management to: Assess the yield reduction due to plant loss Determine the amount of plants which require replacement Define the replanting locations
The total study area had a size of 120ha. The image analysis revealed that over the whole area more than 12,000 banana plants require replanting. In the worst sections of the plantation up to 320 plants per hectare need to be replaced.
Comparing Actual vs. Target Plant density
Instead of mapping gaps and mortality, vital plants can be digitally detected by image inspection and plant densities can be derived. Single plants can be easily identified in an UAV image. Once digitally mapped, they can be automatically counted and plant densities can be calculated.
The fields were flown 2 months after planting. Single plants can be easily identified. Each yellow circle represents a pineapple plant. The average plant density was found to be 28,700 plants/ha. Compared to the target density of 31,000 plants/ha, the actual density is too low.
The plantation management measured the distance between the plants and found that the target plant distance has not been implemented correctly in the field and they were often planted too far away from each other. In effect, the planting crew was either not skilled enough or was not following management instructions: in reality both are prevalent risks in plantation management.
What UAV surveys can give to a plantation management is the full picture, in great detail, of their plantation and control methods that allow intervention and potential for improved profitability, at an early stage of growth. This study shows how both mortality and improved canopy coverage can be detected on a large scale and effective, preventative measures be taken.