Case Study - Precision Agriculture

Signpost Surveys have undertaken detailed research on the use of Unmanned Aerial Systems for Precision Agriculture.

Employing the use of our eBee Ag UAS fitted with a Near Infra Red (NIR) Camera Signpost Surveys have collected multispectral data across a variety of crops.

By analysing the spectral reflectance values of the NIR, Red and Green bands Signpost Surveys can identify problematic areas within a crop. Based on this analysis management zones are identified and nutrient/herbicide/pesticide recommendations are made.

Signpost Surveys products serve to identify which areas within a crop are under performing and why. This information can then help to concentrate your efforts on these areas, thus reducing your inputs and increasing your outputs.

What sets Signpost Surveys products apart from traditional ground based testing methods is that our multispectral imagery can see what the naked eye cannot. This allows us to detect issues that may become problematic at a far earlier stage. In addition the speed of analysis allows for next day results which in turn enable action to be taken when it is needed.

The emergence of GPS technologies and variable-rate machinery have allowed for the development of precision agriculture. Unmanned aerial systems are the latest development and will serve to improve precision agriculture by collecting larger amounts of data in a shorter period of time.

Signpost Surveys now have the ability to help our clients move from sampling a crop to measuring the entire population of plants.

Heretofore nitrogen applications after crop emergence have been based on the average of samples taken within a field or part thereof. However this method may cause too much nitrogen to be applied to half of the crop and too little to the other half. With the emergence of variable-rate applicators, it is possible to adjust nitrogen levels to target particularly problematic areas which will increase yields and reduce costs.

Signpost Surveys offer our clients highly accurate assessments of crop health and vigor. These assessments form the basis of accurate recommendations for fertilizer/herbicides/pesticides. Through the analysis of our imagery Signpost Surveys can identify problematic areas or management zones within a crop. These management zones are then used to apply fertilizer/herbicides/pesticides with far more precision than ever before.

The following case study outlines how these management zones are identified within a crop and highlights how our verified results can help reduce your farm inputs whilst increasing your farm outputs.

The images above show results from our eBee Ag UAS fitted with our NIR camera. The image on the left is a colour composite of all three bands, NIR, Green and Red. The image on the right is a processed Normalized Difference Vegetation Index (NDVI) image. These two images show how detail and information that is not visible to the naked eye and not visible in unprocessed multispectral imagery can be extracted through processing.

Signpost Surveys conducted a complete survey of a crop of Spring Barley in late July 2014, the can crop can be seen in the centre field in both images above and in the index mosaics below. This crop was approximately 6-8 weeks from harvest and at Growth Stage 75 at time of survey. The imagery obtained during this survey has formed the basis of a case study to verify UAS applications in precision agriculture.

In order to identify areas within the Spring Barely crop which have not progressed at the desired rate, produced a reduce yield and are showing signs of stress the raw NIR, Red and Green bands were analysed to produce an NDVI and Green NDVI index mosaic. The Green NDVI mosaic can be seen above.

To obtain a basic understanding of how to analyse and interpret these images a basic knowledge of remote sensing and spectral reflectance is needed. All plants absorb and reflect different portions of the electromagnetic spectrum, by measuring the amount of absorbed and reflected light we can obtain an extremely accurate indication of plant health and vigor.

As we now know all plants absorb and reflect light. A healthy plant will reflect the Near Infra Red and Red portions of the electromagnet spectrum but will absorb the Green and Blue portions. Depending on what growth stage the crop is at we can derive a lot of information about soil nutrients and external stress factors such as fungi (ramularia and net blotch) and aphids long before issues become visible to the naked eye.

As the crop surveyed in this case study had already headed out and was coming into its final stages of growth the two bands that would reveal the most information were the NIR and Green bands. The NIR band was chosen to give a good overall indication of plant health, the Green band would provide a good indication of plant vigour and soil Nitrogen levels. As already explored plants absorb green light and use this energy in photosynthesis to produce chemical energy. This chemical energy is then coupled with the correct levels of Nitrogen for leaf growth and development and Potassium for increased yield and resilience.  A weak spectral reflectance in the Green band suggests poor plant vigour and low levels of nitrogen and potassium.

To verify the results obtained from the NIR imagery and G-NDVI mosaic a number of soil samples were taken from areas which had been identified after analysis as healthy and unhealthy/under stress. The test locations can be seen in the imagery below.

Purple indicates healthy vigorous vegetation, blue stressed vegetation, green unhealthy weak vegetation and red no vegetation.

Soil samples were taken at locations B and C as these areas within the G-NDVI mosaic indicated a healthy and vigorous crop suggesting correct levels of macronutrients Nitrogen Phosphors and Potassium in the soil. Sample sites A and D were chosen as these areas within mosaic indicated the crop was under developed and slightly stressed.

All samples were sent to an independent testing laboratory for pH, macronutrients (P and K) and trace element testing. Separate testing was undertaken for Nitrogen levels.

Also please note the effects of soil compaction in both images. Machinery track lines are clearly visible (orange)around the NE headland of  the top right two fields in Image A/B. In Image C/D Machinery tracks lines and turning circles are clearly visible in the two fields in the lower left of the image. This soil compaction will have adverse growth effects during next years season. Although you may save time during a good dry harvest season, such as 2014, by allowing heavy trucks into the field their excessive weight will only create compaction issue for next year.


Preliminary soil test results have verified the indications within the index mosaics.

Soil sample sites B and C have shown a pH value 6.6/6.5 (very close to the optimum value of 6.6 for spring barley) and higher levels of Nitrogen, Phosphorus and Potassium (N, P, K) soil sample sites A and D have shown a slightly more acidic pH value of 6.7/6.8 and low levels of N, P and K.

As stated these are preliminary results and a full report from the laboratory, including trace elements will be obtained next week. The laboratory report and its findings will be published here once received. This data should conclusively verify the use of Unmanned Aerial Systems for Plant Health/Stress assessments and Yield monitoring.

Full Lab Report

The full soil test lab report including, trace elements, have verified the use of UAS imagery in Precision Agriculture. The results of each soil sample site will be examined in detail and compared against its corresponding Index Mosaic result.

Soil Sample Site A

As previously stated Site A was chose as it gave a weak result in the G- NDVI image, as seen on the right.

The results from the lab have proven that Site A has slightly acidic pH content for spring barley production and is low in N, P and K. In addition test for trace elements (Magnesium, Zinc, Copper and Manganese).

The slightly acidic pH content and lack of macronutrient has impacted on the growth, development and health of the crop in this area.

Original lab report for Soil Sample Site A can be seen below.

Soil Sample Site B

Site B was chose as the G-NDVI image indicated a healthy and vigorous crop suggesting correct levels of nutrients in the soil.

The lab results proved that Site B did indeed have the correct levels of N, P and K and also had the preferred pH value of 6.6.

Having the correct level of nutrients in the soil would most certainty have contributed to the growth and development of the crop.


Analysis of Soil Samples A and B

From reviewing the G-NDVI image it was clear to see that Sites A and B were producing different results. As Site B was producing a stronger spectral reflectance than A it was believed that the crop located at Site B was healthier, in better condition and would therefore produce a greater yield.

The soil samples taken from Sites A and B have now also shown a clear difference in the levels of macronutrients (N,P,K). From these tests we can prove that there is a direct correlation between correct levels of soil nutrients, plant health and development and UAS Index mosaics derived from spectral reflectance values.

In this particular case both sample sites were proven to be low in micronutrients, almost to identical levels, therefore no conclusion on micronutrients can be provided.

To fully corroborate the finding that UAS Index Mosaic results, soil macronutrients and plant health/development are directly related the same results would be required from sample sited C and D.

Soil Sample Site C

As with the previous sample sites, Site C was chosen on the basis that it delivered a strong response in the G-NDVI image, as seen on the left.

The soil test results from Site C revealed almost identical results to Site A. Correct levels of N, P and K were present however pH levels were slightly under the optimal 6.6 at 6.5.

Soil Sample Site D

Sample Site D which will be used to compare against Site C was chosen as it provided a weak return in the G-NDVI image.

The lab results for soil samples taken from Site D were again similar to the results obtained from Site A (weak G-NDVI return). Site D was proven to be low in soil nutrients N, P, K and to have a slightly higher pH value at 6.7.

Analysis of Soil Samples C and D

Once again the clear difference that existed in the G-NDVI index mosaic was replicated in the soil test results from the same locations. The area of crop which was identified by the G-NDVI image as being less healthy/under stress also had the least amount of macronutrients in the soil.

A slight difference was noticed in the levels of micronutrients between Site C and D. Site C which indicted a healthy well developed crop and had high levels of macronutrients also had higher levels of Zinc, although this level was still below normal. Without further testing it is difficult to derive any relationship between UAS Index mosaic, soil micro nutrients and crop health and development.


The results from this study have conclusively proven that UAS index mosaics can be used to identify crop health and stress, identify problematic areas within your crop and allow for an early targeted treatment.

Plants need the correct levels of nutrients in order to thrive and produce a strong yield. The correct levels of Nitrogen will ensure strong growth of vegetation and foliage, correct levels of phosphorous are required for strong root and stem growth and correct levels of potassium is needed to improve resistance to disease and also ensure a better quality crop.

If any of these nutrients are lacking in the soil the plant will become stressed and struggle to thrive.

UAS Index mosaics now offer the ability to identify exactly what areas of your crop and stressed or struggling and to directly target these areas.

UAS NIR/Multispectral imagery can identify these management zones long before the problem become visible to the naked eye. This means that these management zones can be targeted before crop development and yield is affected.


UAS Index Mosaics = Management Zones

Management Zones = Targeted Treatments

Targeted Treatments = Less Inputs with higher outputs