Why speck detection is important for flour millers

Magnus Lindgren,
Perten Instruments AB, Instrumentvägen 29-31, SE-126 53 Hägersten, Sweden
mlindgren@perten.com

Specks, or darker bran particles in flour, have been a specification parameter for durum semolina for quite some time. The specks negatively influence the visual appearance in pasta, and customers perceive the product as poor. In wheat flour milling, the detection of specks is of less concern; it is not a common specification parameter for wheat flour.

From a processing point of view, the detection of specks can help a miller avoid running scrap, customer rejects and costly reworks.

The DA 7300 contains a digital camera and a dedicated illumination system (Figure 1). The camera produces high-quality images of the product that is measured and provides a live video for the operators so they can continuously visually monitor the product in the stream (Figure 2).

DA 7300
Figure 1. DA 7300

Figure 2. Live camera images and video
Figure 2. Live camera images and video for visually monitoring the product stream

Once the images are captured, the DA 7300 performs an image analysis whereby the specks are detected and counted, the size of the specks is noted, and they are grouped according to their size. Although the users typically view only one size that corresponds with their manual reference method, the system identifies 16 different groups (Figure 3).

Figure 3. Specks grouped by size
Figure 3. Specks grouped by size

It is also possible to separate specks into two categories based on their darkness. Note the blue circle indicating a dark speck and the red indicating a light speck particle (Figure 4).

Figure 4. Categories of darkness 
Figure 4. Categories of darkness: blue circle – dark speck; red circle – light speck

While from a quality point of view detecting and counting the specks is of interest, there is one aspect that overshadows the others when milling wheat flour. Tempering is designed to soften the bran particles to ensure they are soft, not brittle, in order to avoid them cracking into small pieces and passing through the sifter. However, a small amount of specks normally appears, as one cannot expect to sift away 100%.

In addition, the woven fabric though which the flour is sifted wears and sometimes breaks. When this happens, the amount of speck particles can increase dramatically in a short time. The miller then needs to stop the milling process and replace the broken cloth. The challenge for the miller though is to know when this happens, and the DA 7300 speck detection provides invaluable information to detect breakages immediately.

Real-life scenario

Please note how the red trend for specks increases in Figure 5.

Figure 5. Snapshot of monitoring curves
Figure 5. Snapshot of monitoring curves

By looking at the camera images (Figure 6), the miller, a careful gentleman, verifies that the specks have indeed increased. It now becomes obvious that the mill needs to be stopped and the cause found.

Figure 6. Snapshot of monitoring images
Figure 6. Snapshot of monitoring images

The frames are pulled from the sifter cabinets and inspected (Figure 7).

Figure 7. Inspection of the frames
Figure 7. Inspection of the frames

A tear is found quickly, with a minimum of sub-specification flour produced (Figure 8).

Figure 8. The tear causing the problem
Figure 8. The tear causing the problem

In this case, the miller was able to detect the specks, shut the milling down, replace the broken sifter fabric and be up and running again within an hour.

Read more about the DA 7300