Research On Automatic Orientation System Of Fructus Aurantii Based On Machine Vision

Fructus aurantii is a kind of chinese medicinal material with the effect of broadening qi and stagnation and reducing swelling.The immature fruit must be cut in half from the middle of the fruit stem-calyx axis and then dried to ensure the superior quality of fructus aurantii.At present,due to the lack of directional splitting technology for fructus aurantii,its processing process relies heavily on manual operations,especially the attitude orientation link in the slicing and drying links.In order to save labor costs and improve the automation level of processing of fructus aurantii,it is urgent to solve the problem of automatic orientation of fructus aurantii.In this paper,the fructus aurantii is used as the research object,combined with machine vision technology and orientation adjustment actuator to develop the fructus aurantii automatic orientation system.The main research contents are as follows:(1)Through the analysis of the shape,surface characteristics and orientation requirements of fructus aurantii,the architecture and automatic orientation process of the fructus aurantii automatic orientation system based on machine vision are designed,and the specific system hardware and software design is completed.(2)Aiming at the problem of segmentation of multiple types of regions of interest in fructus aurantii,a fructus aurantii image segmentation algorithm based on Otsu threshold segmentation and red-green difference method is designed to achieve effective filtering of background and some non-fruit stem area.In order to distinguish the regions of interest at the edge and non-edge positions of fructus aurantii in the image of fructus aurantii,an algorithm for determining the location of the region of interest was proposed,and the color,texture and shape features of the regions of interest at these two different locations were completed 7 and 9 feature value extraction.(3)To solve the problem of fructus aurantii posture recognition,a fructus aurantii posture recognition algorithm based on fruit stem marks was proposed.The SVM classifier is used to realize the classification and recognition of the regions of interest in the edge and non-edge positions,respectively,and the recognition accuracy rate reaches 92.6%and 95.6%.The orientation of fructus aurantii was analyzed using the orientation of the axis of the fruit stem scar,the fructus aurantii model was constructed,and the deflection angle calculation method of the fructus aurantii was determined.In order to solve the problem of using multiple different SVM prediction models to identify multiple fruit stem marks,a decision tree optimization algorithm was proposed.(4)The orientation adjustment actuator is used to adjust the orientation of the fructus aurantii.By analyzing the relationship between the deflection angle of the attitude recognition output and the vertical posture,the optimal orientation adjustment angle of the fructus aurantii in different postures is obtained.Movement planning for fructus aurantii orientation adjustment.(5)Based on the results of the above research work,the automatic orientation system of fructus aurantii based on machine vision was built to measure the clamping force parameters of the fructus aurantii and test the prototype.Experiments show that:under the allowable slight deformation of the fructus aurantii within 0.5mm,the clamping force of 2.45N and 3.37N can be used for the fructus aurantii of 45?55mm and 56-65mm,respectively,to ensure that the fructus aurantii is clamped.The success rate of the orientation of the fructus aurantii is 94.5%,the maximum orientation error after two orientation adjustments is 4.8°,and the average time for a single fructus aurantii is 8.1s,which can meet most of the requirements for the operation of fructus aurantii.