Access Official Technical PDF

Share Summary:LinkedInEmail

Water turbidity is a useful indicator of suspended particles, but conventional measurement equipment can be costly. A smartphone turbidimeter offers a different approach: using a phone camera to detect light scattered by a water sample, then applying image processing to estimate turbidity. A reported low-cost design demonstrated this method with prepared Formazine samples across a defined measurement range.

How a smartphone turbidimeter measures turbidity

Turbidity measurement relies on how particles in water interact with light. In the described setup, a white LED illuminates a sample held in a glass cuvette. Suspended particles scatter some of that light toward the smartphone camera, which acts as the optical detector. The camera records an image, and software converts it into a grayscale representation for analysis.

The grayscale value expresses image brightness on a scale from 0, representing black, to 255, representing white. The method calculates the mean greyscale intensity (MGI) across the image, then relates that value to a reference turbidity measurement. In this design, brighter images were associated with higher turbidity because more particles scattered light toward the camera.

Why the measurement enclosure matters

The camera’s reading depends not only on the sample but also on the conditions under which the image is captured. The design therefore places the LED, cuvette and smartphone in a small cabin made from corrugated cardboard. Its interior is painted black, helping block external light and providing a dark background for the sample.

Two openings help control the optical path. A narrow slit directs LED light toward the sample while limiting direct light reaching the camera. A pinhole allows scattered light to reach the camera from approximately a 90-degree direction and helps reduce stray light. The enclosure is a low-cost alternative to more elaborate optical components, although it also means that repeatable positioning and consistent setup are important when taking measurements.

The light source is a white LED powered by two AA batteries. White light was selected to cover the visible spectrum, while the LED’s affordability and simple power requirements suit a compact, inexpensive instrument. The smartphone used in the study captured images with automatic camera sensitivity settings. Images were saved at a reduced resolution to make processing faster.

Calibration and reported measurement range

An image-based sensor needs calibration before its output can be interpreted as a turbidity value. The design was calibrated using 33 prepared Formazine concentrations spanning 0 to 100 nephelometric turbidity units (NTU). Each sample’s reference turbidity was measured with a Hach 2100P turbidimeter, and its image-derived MGI was compared with that reference.

The resulting linear calibration relationship was reported as turbidity in NTU equal to 11.705 multiplied by MGI, with a coefficient of determination (R²) of 0.982. R² indicates how closely the calibration model fits the measurements used to develop it; it does not, by itself, describe every source of measurement uncertainty. The result supports the feasibility of estimating turbidity from images under the study’s test conditions.

The demonstrated range is limited to 0–100 NTU and the calibration samples were prepared Formazine. The reported result should therefore be understood within those conditions, rather than as evidence that the same calibration will apply automatically to every water source or measurement setup. Samples, camera settings, illumination and enclosure geometry all form part of the measurement system.

Practical considerations for industrial and field use

For automation engineers and water-monitoring teams, the key lesson is that a camera-based measurement is a complete sensing chain, not simply an image of a sample. The chain includes controlled illumination, sample presentation, camera capture, grayscale conversion and a calibration that maps image data to NTU.

  • Control ambient light: An opaque enclosure and consistent lid position help limit external-light interference.
  • Keep geometry consistent: The LED, cuvette and camera need repeatable positions for comparable readings.
  • Use an appropriate calibration: The demonstrated model was developed for Formazine samples over 0–100 NTU; other applications require suitable validation.
  • Manage sample timing: Measurements in the study were taken soon after preparation because the Formazine samples were stable only for a short period.

The reported fabrication cost was USD 4.35, excluding the smartphone and cuvette, which were already available to the study. That figure reflects the specific materials and components used, including repurposed cardboard, a white LED, batteries and paint. It is not a complete cost for an installed or production-ready monitoring system.

What the result means for water monitoring

A smartphone camera can be used as a light detector in a basic turbidity instrument when the optical setup is controlled and the image signal is calibrated against reference measurements. The reported R² of 0.982 across 0–100 NTU shows a strong fit for the prepared Formazine samples tested, while the limited range and sample conditions define how broadly the result should be applied.

For practical deployment, the approach is best treated as a foundation for an inexpensive measurement concept rather than a universal replacement for professional instruments. Teams considering it should account for the required calibration, repeatable sample handling and controlled lighting. Those fundamentals determine whether an image-based reading can be useful in a particular water-monitoring workflow.

Access Official Technical PDF

About The Author