f2m Automation Book

Page 97

IMAGE PROCESSING

97

Image processing applications for baking process monitoring Computer vision and image processing offer the development and employment of non-invasive, contactless, non-destructive, and fast measurement methods with a wide range of applications within the baking industry. These include both the visualization (qualitative measurement) and the exploration of shape and

+

As a subcategory of signal processing, image processing is used in satellite imaging, medical imaging, and industrial imaging for process and even product assessment. Image processing uses camera technologies to capture two- or three-dimensional data, followed by data transfer to a processing unit to extract product or process information. The usual captured light range is the visible (VIS) light range (400–800nm) because it is captured by both CMOS (complementary metal-oxidesemiconductor) and CCD (charge-coupled device) cameras, which can also include the initial nearinfrared (NIR) range (800–900nm). Silicon sensors can also detect the extended VIS/NIR range. The NIR range up to 1700nm is captured using InGaAs sensors, e.g., with shortwave infrared (SWIR) cameras. For the mid-infrared range (MIR, 2.5–25.0μm), used to capture thermograms, microbolometers are utilized. Microbolometers can also be used to capture ultraviolet and Xradiation. Other camera types for specific parts of the electromagnetic spectrum are available, and these are constantly being developed and adapted to match changing requirements. Image data capture occurs mainly in line scan mode, in which a dynamic scene is captured line

by line, or in area scan mode, where the whole scene is captured in one shot. For image processing, these single recorded lines are sequentially stitched together for image reconstruction. For example, with conveyer belts, typically line scan cameras are used to capture continuous webs of raw material or products. The obtained image data are available as twodimensional matrices with intensity values. For monochrome cameras, the form is n×m×1, and for RGB cameras, the form is n×m×3. The light or color intensity forms the third dimension for each pixel inside the layers in the n×m matrix. InGaAs-cameras and microbolometers capture n×m×1 monochrome data. By using filters for the separation into individual wavelengths, so-called spectral observations can be carried out. In the resulting spectral cube, the captured data is in the form n×m× ‘number of observed wavelengths/ regions’. The number of captured image data (sampling frequency) is specified by sampling criteria and indicated by frames per second (fps). Image data processing can be applied at different computational levels. First is the description of the image using statistical tools, e.g., the mean, minimum and maximum value of the color, contrast, or histogram analysis. The image data

I M AG E P RO C E SS I N G A P P L I C AT I O N S F O R BA K I N G P RO C E SS MO N I TO R I N G

surface properties (quantitative measurement).


Turn static files into dynamic content formats.

Create a flipbook

Articles inside

WP BAKERYGROUP: Connected processes

9min
pages 175-178

TECNOPOOL S.p.A.: Complete spiral system control

3min
pages 173-174

Rademaker B.V.: Training is money well spent

9min
pages 167-170

Sugden: Baking for joy

2min
pages 171-172

MECATHERM: The human must remain the pilot

8min
pages 163-166

Koenig Group Baking Equipment: The future of the baking industry is automation

4min
pages 161-162

Kaak: Bring time on your side

9min
pages 157-160

Heuft Industry: Energy savings at the end of the tunnel oven

8min
pages 153-156

FRITSCH Group: Progress in the world of bakery

11min
pages 149-152

Diosna: Everything from a single source

4min
pages 143-144

Ernst Böcker: Why sourdough plays a decisive role

6min
pages 145-148

Cetravac: Fast, flexible and sustainable

4min
pages 141-142

AMF Bakery Systems: Future-smart technology arrives

11min
pages 135-138

Bakon: The key is knowledge

4min
pages 139-140

American Pan: Pan design and handling for automated bakery systems

7min
pages 131-134

Cybersecurity: Safe and smart bakery production

8min
pages 123-130

3D printing: Will we 3D print the bread of the future?

26min
pages 113-122

Artifical intelligence: The role of artificial intelligence in designing baking ovens

12min
pages 105-112

Image processing: Image processing applications for baking process monitoring

15min
pages 97-104

Design thinking: Using design thinking to facilitate automation

22min
pages 87-96

Digitization: Digitizing food supply chains

15min
pages 79-86

Smart stores: The search for answers is on

20min
pages 23-32

Rheology: Bread dough rheology

17min
pages 33-40

Mixing: Dough mixing supervision: an overview

21min
pages 51-60

Baking line audit: Metrology on baking and freezing lines

25min
pages 41-50

Robotics: Autonomous performance

12min
pages 17-22

Software: Manufacturing Execution Systems in bakeries

17min
pages 9-16

Digital twins: Digital twins in baking process automation

14min
pages 71-78
Issuu converts static files into: digital portfolios, online yearbooks, online catalogs, digital photo albums and more. Sign up and create your flipbook.