Assessing and detection of multiple bruises in peaches based on structured hyperspectral imaging

Area of Science:

• Agricultural Engineering
• Food Science
• Optical Engineering

Background:

• Postharvest damage, such as bruising from impact, falling, or compression, significantly reduces peach quality and marketability.
• Accurate and non-destructive detection of bruising is crucial for quality control in the fruit industry.

Purpose of the Study:

• To investigate the efficacy of structured hyperspectral imaging (S-HSI) for detecting various types of postharvest damage in peaches.
• To develop and evaluate bruise detection models using S-HSI data and chemometric methods.

Main Methods:

• Acquisition of S-HSI images at spatial frequencies of 60, 100, and 150 m⁻¹ with a 2π/3 phase shift interval.
• Processing of S-HSI data using a mathematical demodulated model to generate high-resolution image cubes.
• Development of bruise detection models using artificial neural networks and principal component analysis on S-HSI spectra.

Main Results:

• S-HSI spectra demonstrated superior discrimination capabilities for bruised versus healthy peach samples compared to ordinary hyperspectral spectra.
• Optimal bruise detection models were achieved using combined spatial frequencies of 100 m⁻¹ and 150 m⁻¹.
• The models successfully differentiated between healthy peaches and those with impact, falling, or compression damage.

Conclusions:

• Structured hyperspectral imaging is a highly effective non-destructive optical technique for detecting postharvest bruising in peaches.
• S-HSI offers a promising approach for quality assessment and damage detection in horticultural products.
• Further research can explore S-HSI for real-time quality control systems in the food industry.