Research On Citrus Internal Quality Detection Based On X-ray Imaging Technology

Citrus is the largest type of fruit in the world,which is rich in resources and has a great variety.It has a high nutritional value and health value.Consumers all over the world love its unique color,aroma and taste.Due to the stress of the growing and storage environment of citrus,physiological diseases,such as peel puffing,granulation and mildew,will appear inside the citrus,which will seriously affect the quality of citrus fruit and cause the loss of market value of citrus.Traditional detection methods of citrus internal quality mainly focus on the detection of physiological and biochemical indexes,which not only have a large workload,but also have damage to citrus and low efficiency,so it is impossible to achieve mass detection.In this paper,X-ray imaging technology was used to study the peel puffing and granulation of citrus:1.The X-ray imaging detection device was built,and the detection parameters were determined according to the physical properties of the tangerine.Taking the clarity,contrast and distortion rate of citrus image as evaluation criteria,the best detection parameters of X-ray imaging device were determined as follows: tube voltage of X-ray source was 60 k V,tube current was 1.3 m A,integral time of linear array detector was 5.5 ms,and transmission speed of orange conveyor belt was 10cm/s.2.The geometric distortion and distortion coefficient of X-ray imaging were calibrated.X-ray imaging distortion mainly includes column direction distortion and row direction distortion.The sudden change of transmission speed will cause the image distortion in the column direction.Under the influence of X-ray fan-shaped beam irradiation,the image distorts in different degrees from the center to both sides in the row direction.The metal plate with the same hole spacing was used as the calibration plate to detect the distortion in the column direction.The detection results show that the hole spacing is consistent with the calibration plate,which indicates that the transmission speed is stable,that is,the distortion in the column direction can be ignored.A 70 mm stainless steel calibration ball was used to detect the distortion in the row direction.The projection distortion coefficient of different positions in the row direction was calculated and the distortion model was established.3.Citrus image was preprocesssed.3*3 Gaussian filter was used to denoise citrus image.Image enhancement algorithm is was used to enhance the contrast of the image after denoising.The fixed threshold segmentation method was fused with the morphological algorithm to remove the background region of the citrus image and separate the citrus pulp region and peel region.4.X-ray imaging technology was used to calculate the transverse diameter of citrus fruits.The pixel accuracy of the image was calibrated with 60 mm,70mm and80 mm stainless steel balls,and the pixel accuracy was 0.35mm/pixel.It was achieved the average error of the transverse diameter of citrus was 2.45 mm,and the maximum error was 6.67 mm by X-ray imaging technology.5.Citrus quality characteristics were extracted.The pulp area and fruit area ratio of citrus were calculated to determine the degree of peel inside the citrus.The discrimination accuracy rate of this method was as high as 96.2%,and the misjudgment rate was 3.8%.The gray characteristics of citrus fruit region were extracted,and the ratio of citrus dry area to citrus pulp area was calculated to distinguish the degree of citrus granulation.The accuracy was 86.9%.It was found that the larger the citrus was,the more serious the internal peel puffing and granulation were.The peel puffing index and granulation index of L-fruits(70-80mm)were 54.6 and 41.7,and that of M-fruits(65-70mm)were 38.6 and 31.8,respectively.and that of S-fruits(60-65mm)were 34.1 and 26.5,respectively.