Methods of rapid bruise assessment and the formulation of robust bruise indices for potatoes
Evans, Stephen David
When potato tubers are subjected to impacts, the sub-surface tissue may become discoloured as damaged cells produce the blue-black pigment melanin. Bruising caused during harvesting and handling can lead to downgrading of potatoes for the processing industry and quality retail trade. The two aims of this thesis were to reduce the time to detect bruising, and to develop a non-subjective method for the quantification of bruising.Reflectance spectrophotometry was investigated as a rapid, non-subjective and non- invasive way of detecting bruising. Wavelengths from ultraviolet to near infrared were selected by discriminant analysis to separate unbruised and bruised tubers. Neural nets were trained with these wavelengths to identify bruised tubers in a sample of unbruised and bruised tubers. The detection of bruising gave inconsistent results in unpeeled tubers, but proved to be reliable in peeled tubers.The rate of bruise development at air pressures up to 10 bar was measured by reflectance spectrophotometry and by a visual rating. The production of dopachrome, an orange pre-cursor pigment of melanin, was used as an early indication of bruising. Dopachrome is visible to the human eye and the time for bruise detection can be reduced to approximately 3 hours when compressed air is used.Infrared and microwave thermography were used to measure possible rises in bruised tissue temperature. Thermography was used in conjunction with scanning laser Doppler imaging to detect changes in the biological zero of bruised tissue. No significant differences could be detected between unbruised and bruised tubers using these techniques.Reflectance spectrophotometry was also used in combination with a colour digital camera to automatically measure bruise area in peeled tubers. While the camera alone could measure bruise area with precision, constant adjustments were needed. Reflectance spectrophotometry was faster but less precise than the camera for measuring bruise area.A new bruise index was formulated by taking measurements of the volume of a tuber removed per peeler stroke for the variety Record, and using the data to model the percentage volume lost for any variety. By counting the number of peeler strokes to remove a bruise an estimate of the tuber volume lost could be derived. This method can accommodate any type and size of bruise and results in an index that can be easily interpreted.