Quality inspection techniques of fruits and vegetables using hyperspectral imaging: A review

The Hyperspectral Imaging concept is widely used in remote sensing, but with the advances in semiconductor technology and modern times, this concept is nowadays used for the quality inspection of fruits and vegetables. It provides a non-destructive and fast way to assess internal and external quality parameters, helping to improve safety, quality, and processing efficiency. The importance of fruits and vegetables is vast and diverse in global food systems. However, post-harvest losses due to inefficient processing methods significantly impact farmers' income, food security, and industry profitability. Advanced automation and precision technologies have the potential to revolutionize unit operations in post-harvest processing, making them more efficient, scalable, and sustainable. The Hyperspectral Camera captures many images at specific wavelengths, far beyond the visible range. These images not only provide information about the visible features like color but also about the chemical and structural composition of the fruits and vegetables. Hyperspectral data often requires spectral correction to account for various environmental factors, like lighting conditions or sensor calibration errors as a part of pre-processing. The selection of wavelengths provides useful information for quality assessment. Image processing techniques are applied to enhance these images, such as filtering, noise reduction, extracting features, and segmenting regions of interest to isolate specific areas of the image that have indicated potential defects or abnormalities. In this way large amount of data is collected and the next step involves extracting meaningful patterns from the data generated by hyperspectral imaging. Then, machine learning algorithms are applied to reduce and classify this data to a manageable size and identify key features that correlate with the quality characteristics of the fruits. After processing, the data is analyzed to produce a quality score for each fruit or vegetable.