Hyperspectral image reconstruction method and detection method for fruit quality detection

• Main Authors: ZHU GONGQIN, LIU CUNCHUAN, WENG SHIZHUANG
• Format: Patent
• Language: Chinese; English
• Published: 21.09.2021
• Subjects: INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIRCHEMICAL OR PHYSICAL PROPERTIES; MEASURING; PHYSICS; TESTING

The invention particularly relates to a hyperspectral image reconstruction method for fruit quality detection. The method comprises the following steps: S110, shooting a fruit to obtain a high-definition image and a hyperspectral image of the fruit; S120, segmenting the two and extracting a region of interest; S130, averaging the camera channel response value D and the spectral reflectivity R to obtain a sample set; S140, finding a conversion matrix W according to the sample set; and S150, shooting a high-definition image of the to-be-detected fruit by using a high-definition camera, and converting the channel response value Dv of the to-be-detected fruit by using the conversion matrix W to obtain reconstructed spectral reflectivity data Rv. The invention further discloses a method for fruit quality detection. A hyperspectral image reconstruction technology is adopted so that the instrument and equipment cost and the manual operation time cost are greatly reduced; the operation and application are simple, the