Labwise: Dynamic Image Analysis

If one wants to, not only determine the size of the particles in powdered materials, but also learn something about their shape, dynamic image analysis is a method of choice. It facilitates taking measurements in both dry, free-flowing bulk solids and in a wet dispersion unit.
The underlying measuring principle is simple. A stream of particles is passed in front of a large scale LED flash and the particles are photographed in the back light of a digital camera. This results in a high contrast between the light background and the light-shielding particles. The images are then analyzed by software.

The Range of Particle Size

The size of the image produced on the camera sensor depends on the magnifying power of the lens being used. The limits of the size range that can be recorded with a certain combination of the camera and lens are obtained from the magnification and the sensor parameters (total size and pixel size). For example - a 5 megapixel camera with a 2/3-inch CMOS sensor and a pixel size of 3.45 μm in combination with a lens magnification of 0.35x results in an object size of about 10 μm per pixel. With particle images of at least 8 x 8 pixels, a lower measurement range of 80 μm and an upper measurement range of around 10 mm can be obtained.

When using a fixed magnification factor, this is only possible with a regular camera lens if the distance between camera and particles is always identical. If the distance changes, the detected particle size is also falsified. Since, in practice, it is almost impossible to direct the particles past the camera in exactly one plane, tele-centric lenses can be used to correct any errors. In this case the size of the image generated on the sensor does not depend on the distance between object and camera.

How are the Particles Recognized?

The software detects dark areas as particles and lighter particles as background. There are numerous gradations between light and dark. The number of different levels of grey available using a camera with a dynamic range of 8 Bits is 256. Absolute white corresponds to a value of 255, black is 0. A threshold is specified in the software indicating whether a pixel belongs to the background or to a particle.

This threshold can be adjusted individually.

Depth of Field

This describes the range of distance within which a particle is sufficiently sharply depicted. A lens’s depth of field decreases with increasing magnification. As a result, the edges of particles which do not pass across the exact focus plane of the camera show a gradual transition from black to white. The software can use this transition to decide which particles are still sufficiently depicted to be used for evaluation.
The image acquisition rate, usually expressed in “frames per second” (fps), is also relevant. At high image recording rates (e.g. 75 fps), large amounts of data are generated within a very short time, which make corresponding demands on the computer hardware. However, this can be reduced if one does not save all the images acquired during measuring and used for determining the results. However, one does lose the basic advantage of the process i.e. being able to view each individual particle.

Wet or Dry?

The quantity of samples needed depends on the particular sample itself and the question linked to the measurement. For most tasks, a few tens to several hundred thousand particles are sufficient, and for large particles even fewer. The quantity of samples needed also depends on whether one is to carry out a dry or wet measurement.

During dry measurement, the sample material is fed into the measurement continuously, whereby the sample feed rate can only be increased within narrow limits. The overlapping of two particle images which randomly pass on the same visual axis of the camera should be kept as small as possible. With this method, each sample particle is only available for analysis once and, especially in the case of samples with a wide range in particle size, the entire sample quantity introduced into the conveyor system should be processed at once in order to avoid influencing the result due to possible segregation en route through the system. Using the dry measurement, particles of about 20 μm to 20 mm can be measured. In practice, it is important to consider how much material is necessary for statistically reliable measurements.  

In the case of wet measurement, it is usually even more important to ensure that the sample is well distributed if one wants to produce a representative sample. A closed liquid circuit is pumped continuously through a measuring cell. The required sample quantity is usually much lower than for dry measurement. For wet measurement, the upper measuring limit is determined by the geometry of the measuring cell. In modern devices, one can measure particles ranging in size from one micrometer to one millimeter.


The particle sizes can be determined from the obtained images. While static light scattering follows the usually incorrect assumption that each particle is spherical, thereby obtaining only one value for the particle diameter, a dynamic imaging method offers several possibilities to define the diameter of a mostly irregularly shaped particle. An example of this would be the “area equivalent diameter”, the diameter calculated from the particle circumference or the so-called Feret diameter, in which two parallel lines are applied to opposing sides of a particle in such a way that they both touch the particle but do not cross the particle edge at any point.

Besides simply determining the diameter of a particle, a major advantage of Dynamic Image Analysis lies in the possibility of obtaining information on its geometry too. One of the simplest shape parameters is the aspect ratio, which is expressed as the quotient calculated from the minimum to the maximum Feret diameter.

So it is possible to generate the distribution and correlation of almost any desired combination of particle parameters. Whether this is a simple size distribution or the relationship between the particle size and the aspect ratio. In a cloud representation, such correlations can be displayed graphically. Each analyzed particle is represented here as a point whose coordinates in the cloud depend on the values ​​of the parameters selected.

Dr. Günther Crolly
Fritsch GmbH
Idar-Oberstein, Germany


Fritsch GmbH
Industriestraße 8
55743 Idar-Oberstein
Phone: +49 6784 700
Telefax: +49 6784 7011

Register now!

The latest information directly via newsletter.

To prevent automated spam submissions leave this field empty.