Non-destructive detection of defects in mango fruit by thermal imaging
- Publication Type
- Contribution to conference
- Authors
- Schulze, K; Nagle, M; Mahayothee, B; Sardsud, V; Müller, J
- Year of publication
- 2012
- Page (from - to)
- 5
- Conference name
- 7th International CIGR Technical Symposium “Innovating the Food Value Chain”
- Conference location
- Stellenbosch, South Africa
- Conference date
- 25 - 29 November 2012
- Keywords
- Mango
The mango variety ‘Nam Dokmai’ is widely produced in Thailand for export, where current quality assessment is done subjectively by visual inspection. There is a need for non-destructive sensing technologies for fruit sorting to realize a more controlled and certifiable product. Advanced sorting technologies are increasingly important for produce from developing countries to meet the standards of high-end international markets.
Interestingly, defects that are undetectable in the visible spectrum can often become visualized via infrared radiation. Infrared thermography has already been investigated for defect detection in some fruits, eg. in apples (Varith et al., 2003; Veraverbeke et al., 2006; Baranowski et al., 2009). Previous investigations found that bruised areas lose water and thus changing thermal properties (Varith et al., 2003). Changes in thermal diffusivity cause a surface temperature contrast, which can be detected by active thermography. Thus, it was the aim of this study to test an infrared thermographic system (spectral range 7.5 - 13 µm) under laboratory conditions (constant temperature at 26 °C) to investigate surface temperature behavior of bruised mangos.
Experiments were conducted with Nam Dokmai fruits subjected to bruising induction with different penetration depths due to ripening stage at bruising date. More than 500 fruits were monitored over ripening during subsequent bruise development. The acquisition was performed as in pulse thermography (PT) with an excitation of a rectangular heat pulse with exposure time set at 30-80 s per sequence, where the surface temperature changes accordingly (Fig. 1). During the exposure time, infrared images were recorded and resulted in 450-1200 images per video sequence at a frame rate of 15 Hz.