Positioning And Identification In The Sorting Process Of Multi-color Waste Beverage Bottles

With the increasing pressure of environment and resources,more and more attention has been paid to the recycling of waste beverage bottles.The sorting of recycled plastic bottles based on their color is a key step in their recycling process.At present,the color sorting of recycled plastic bottles in China mainly depends on manual recognition,but the inefficiency and high labor cost limit the development of the industry to a certain extent.The color sorting device for plastic bottles based on machine vision can greatly improve the efficiency of sorting.This paper focuses on the positioning and color recognition of recycled plastic bottles in the process of automatic classification based on machine vision.The recycling process of recycled plastic bottles is analyzed.And the key equipment used in the recycling process is elaborated.Three automatic sorting schemes for recycled plastic bottles are compared and studied.Based on the image of recycled plastic bottles,the positioning method of plastic bottles is studied.To classify the plastic bottles on the conveyor belt,their position relationships are firstly defined as three categories,i.e.disjoint,adjacent and overlapping.And the method for separating the adhesive bottles is studied.The acquisition process of the image of the bottom of each bottle is described and the color features of recycled plastic bottles are extracted at the bottom of the bottle.After obtaining a large number of data,both the imprecise classification and the precise classification of colors for the plastic bottles are studied in different color spaces.At the same time,the effect of sample size on the color recognition of plastic bottles is also studied.The effect of the layout of light source on the recognition of plastic bottles is studied by combining simulation with experiment.At the same time,the influence of light source brightness and the color of conveyor belt on the recognition of plastic bottles are experimentally studied.The visual system is calibrated.In addition,the treatment method of image sequences of plastic bottles is studied.Finally,the feasibility of the beverage bottle sorting scheme is verified by experiments.