6. CoSpectroCam (CSC) sensor

Written By Nanno Mulder and Ali Abkar

CoSpectroCam (CSC) is an optical coaxial assembly of an RGB Camera and a Spectrometer. The Field of View (FOV) of the spectrometer is mapped as a rectangle onto the RGB image making a link between the spatial features in RGB and the spectral features derived from the reflectance spectra. The FOV of the spectrometer is called “Slit Rectangle”. The development of its implementations are ranging from Indoor, Ground-based to Drone-based systems for different applications as well as education and training in remote sensing or environmental engineering. The following list illustrates the advantages of using CSC for hyperspectral and spatial data acquisition: the FOV of the spectrometer is mapped to the RGB image and displayed on the monitoring screen; RGB images are used to guide the FOV of the spectrometer to optimal view angles and record the RGB image together with the spectrum; spectral data acquisition from any distance to the location of the object (for example, from the ground looking at the sky or clouds); and spatial features such as homogenous areas or texture features are stored along with spectral features in one record identified by the timestamp. CSC comes with software with a GUI for data acquisition, displaying, monitoring and storage of RGB & hyperspectral data. The data are stored as images and Spectra tables with a common timestamp as reference. The Spectra data are stored in table.csv format and hence easily accessible for further use and analysis. The software also includes options for taking DarkObject, WhiteObject, and Object measurements of radiance and reflectance. This agricultural IoT system is enabling experts to produce real-time reflectance data from objects/crops for feature extraction by combining high-resolution spatial features with high-resolution spectral features.

What is Coaxial SpectroCam (CSC)?

PCoSpectroCam (CSC) is an optical coaxial assembly of an RGB Camera and a Spectrometer.

The Field of View (FOV) of the spectrometer is mapped as a rectangle onto the RGB image making a link between the spatial features in RGB and the spectral features derived from the reflectance spectra. The FOV of the spectrometer is called “Slit Rectangle”. The development of its implementations are ranging from Indoor, Ground-based to Drone-based systems for different applications as well as education and training in remote sensing or environmental engineering.

The following list illustrates the advantages of using CSC for hyperspectral and spatial data acquisition: 

• the FOV of the spectrometer is mapped to the RGB image and displayed on the monitoring screen;

• RGB images are used to guide the FOV of the spectrometer to optimal view angles and record the RGB image together with the spectrum;

• spectral data acquisition from any distance to the location of the object (for example, from the ground looking at the sky or clouds);

• spatial features such as homogenous areas or texture features are stored along with spectral features in one record identified by the timestamp;

• CSC comes with software with a GUI for data acquisition, displaying, monitoring and storage of RGB & hyperspectral data;

• The data are stored as images and Spectra tables with a common timestamp as reference.

• The Spectra data are stored in table.csv format and hence easily accessible for further use and analysis.

• The software also includes options for taking DarkObject, WhiteObject, and Object measurements of radiance and reflectance;

• This agricultural IoT system is enabling expert to produce real-time reflectance data from objects/crops for feature extraction by combining high-resolution spatial features with high-resolution spectral features.

The spectrometer catches subtle differences in 2048 channels, such as changes due to fertiliser concentration, environmental stress factors or pests. The management model is based on following plant growth processes through the whole season using the satellite and airborne systems in combination with field sensors. Most important are patches of pilot plants of which the details are known and updated precisely. The pilot plants are used as calibration references for the whole area under management. The Management System is a learning system that adapts to local conditions and management actions. The system can be supplemented by data on soils, climate, and pest risk alerts.

Developing a hybrid sensor system (Called CoSpectroCam (CSC) that combines a hyperspectral resolution of 1024 or optional 2048 channels to a subset of some 200 pixels in the centre of an RGB image for lab, field, and UAV reflectance data measurements. Meeting the challenge to link the scientific concepts of multi-sensor derived information to the management requirements of, for example, farmers.