Farming has developed extensively since the advent of the Agrarian Revolution. New farming methods and equipment have been invented which has made farming easier and more profitable. The revolution experienced in farming progressed in various stages from hand labor, animal labor, farm machinery, satellite navigation to drones. This has resulted in farming practices that use fewer resources and produce maximum yields. It has also brought about advanced methods of crop pest detection.
Crop pests are animals or plants that harm target crops in the farms. Some of the most common pests are insects, bacteria, and fungi. They are a significant threat and contribute to over 30-50% reduction in farm yield. In order to prevent pests from attacking your farm, you need to be aware of their presence before they attack. With recent advancements in technology, you can employ sensors in the field so that they can detect and monitor pest infestation. This article provides insights on how sensors can be used in pest detection.
Sensors Used in Crop Pest Detection
In precision farming, pest detection sensors are instruments used to detect the presence of pests in a farm. They are grouped according to their mode of operation, energy consumption and the type of pests they sense. Some have high resolution and can detect very minute pest organisms hidden deep inside a crop while others only capture images that can be seen by the naked eye.
Spectral Remote Sensors
Spectral remote sensors display data in terms of images. They are divided into low-image sensors and high-image sensors (imaging spectrometers).
- Low-image Sensors
These are cameras that are used on a day to day applications. They are mounted on a trap crop where they capture images and send them to a control station. They can only record images to the extent of their image resolution and since they are low-resolution cameras, they only capture pest images that are visible to the eye. Using the images, you can calculate the number of pests in a crop and also estimate pest infestation on a farm.
Farmers prefer low-image sensors due to their affordability in terms of capital and maintenance. The real-time low-image sensors are also very mobile and can be used to detect pests in a large area within a few days.
- High-image Sensors
All plants and soil reflect a certain amount of light energy to the atmosphere. The amount of light energy they reflect is referred to as the spectral signature. High-image sensors pick up the spectral signature of each crop and record it. These images are recorded in a spectrum beyond the human spectrum. This includes gamma rays, ultraviolet rays, x-rays and infrared. The image data can be produced as multispectral (more than 3 bands) or hyperspectral (hundreds of bands). These high-resolution images can tell the physical and chemical make-up of the plant from hundreds of kilometers away.
In crop pest detection, the imaging spectrometers have pre-recorded spectral signatures of each crop in the farm. Once pests invade a crop, their spectral signature changes since the pests absorb the crop’s light thus forcing them to reflect a different spectral signature. Analyzing the images, scientists can tell the number of pests in the crop, their exact location and the lifecycle stage of the pest. They are better than low-image sensors since they are more accurate and can detect multiple types of pests and diseases. The disadvantage is in their cost. Imaging spectrometers are expensive and require huge resources to maintain them.
Fluorescence Image Sensing
This method involves measuring the amount of chlorophyll in a plant. Images of the crop leaves are recorded and compared with images of a healthy leaf. If there are changes in the chlorophyll pattern, that could be an indication of crop pests. This method can only be used to detect pests where chlorophyll is present.
Sound Detection Sensors
Crop pests can also be detected using sound sensors. These are wireless sensors placed in strategic points across the farm. They have antennas that pick up sound waves in the vicinity. The sensors pick up sound generated by pests when chewing, flying or mating. The farmer then records the noise levels of that place for a certain period. After a short period, the farmer inputs the data into a computer so as to analyze it. Places with pest infestations usually have higher sound waves than the rest. The farmer can choose to control the pests right away or wait till they achieve the economic threshold of control.
It is an affordable crop detection method that can be implemented by large scale and small scale farmers alike. It has little maintenance cost and above average accuracy. The only downside is the influence of environmental factors. In cases of rains or strong winds, you cannot collect data.
Gas Detection Sensors
Plants produce specific chemical volatile compounds when stressed. The stress can emanate from pest infection, environmental factors or human influence. If a pest attacks a crop, the crop releases volatile chemical compounds into the atmosphere which can be recorded by images, or collected as gas. In order to identify the type of pest infestation in the crop, you need to have previous samples of the volatile chemical compounds released by plants after different stresses. You can also use thermal imaging to identify the type of volatile chemical compounds released since they have a specific spectral signature.
Crop pests are dangerous and can cause heavy losses to a farm in a matter of days. Early pest detection using remote sensing can save you money and much stress. Numerous benefits such as accuracy, swiftness and efficiency among others, outweigh the disadvantages of high implementation costs. The amount of work high-image sensors can handle cannot be compared to the traditional methods of crop detection. Also, most crop detection methods do not require the presence of the farmer, thus reducing the cost even further. With remote sensing in crop detection, sustainable farming is secure.
(source: https://www.farmmanagement.pro/farming-revolution-the-use-of-sensors-in-crop-pest-detection/ )