A digital sensor with non-imaging multi-spectral and image modules for continuous monitoring of plant growth conditions: Development and validation

•A low-cost and ground-based digital device coupled with image and non-imaging multi-spectral was proposed and developed.•Multi-spectral module of the designed device showed good performance regarding to light variations in both sunny and cloudy conditions.•Cross-calibration of multi-spectral module exhibited satisfactory results with an ASD 4 FieldSpec spectrometer using paddy rice canopy as observation objects.•Vegetation indices calculated from red-edge wavebands of the developed device behaved with superior results for paddy rice leaf chlorophyll content estimation. In the field of precision agriculture, crop growth monitoring is integral for acquiring a holistic understanding of plant growth status, facilitating the implementation of precise and timely agronomic interventions. While traditional methods rely on labor-intensive sampling surveys, the advent of remote sensing technologies, characterized by multi- or hyperspectral imaging systems, has provided a leap forward in the non-destructive assessment of crop growth parameters. Restrictions due to technical constraints, however, curtail the widespread application of these spectral imaging modalities for plant growth surveillance. To address this challenge, this study endeavored to conceptualize and contrive a cost-effective, ground-based digital apparatus. This instrument, amalgamating a non-imaging multi-spectral module with bandwidths of 10 nm across six distinct wavelengths (530 nm, 570 nm, 680 nm, 700 nm, 740 nm, and 780 nm), a digital camera module, an operational control unit, and a solar cell assembly, was designed for continuous monitoring of crop growth. This was accomplished through intricate hardware engineering and sophisticated system control programming. Performance of multi-spectral module from the designed device were comprehensively evaluated. Results showed that measurements procured from the multi-spectral module were in strong linear correlation with irradiance observations from a canonical irradiance meter (ST-85) across all spectral bands under diverse indoor lighting scenarios, registering non-linear error (NLE) values below 3.1 %, and return error (RE) indices not exceeding 0.7 %. Continuous monitoring under variable outdoor ambient conditions—sunny and cloudy—underscored the multi-spectral module’s commendable performance. Notably, in sunlit conditions, results were particularly promising: statistical metrics of relative error for relative reflectance of each wavelength, alo