Research On Construction Waste Brick-concrete Separation Based On Machine Vision

With the development of modernization construction,the construction waste such as masonry and concrete remaining in demolition and construction continues increasing.Due to the raising cost of labor force,the achievement of automatic sorting of construction waste is of paramount importance.At this stage,the automatic sorting of brick and concrete waste mainly uses a color sorter to spray out masonry and concrete to realizes the separation of masonry and concrete.Due to the problem such as poor working conditions and fluctuations in motor speed,the color sorting machine has a lower accuracy,and making it difficult to achieve accurate sorting.In terms of this issue,this paper designs a QT-based air jet sorting system,which can effectively overcome the disadvantages of low precision,low efficiency and high energy consumption of the current color sorter.The main research contents are as follows:(1)In response to the issues of low brick-concrete detection accuracy and poor real-time performance in harsh working conditions,a conveyor belt target detection algorithm under pollution conditions is proposed.The algorithm utilizes the basic idea of background subtraction,obtains the background image of the current frame image by dynamically locating the position of the current frame image in the overall background image,and solves the problem that the background model cannot be updated in time when the background changes rapidly in the traditional background clipping method.To achieve the detection under the condition of conveyor belt has been polluted.After binarization in HSV space,the classification of brick and concrete is realized.After experiments,the algorithm can effectively eliminate the influence of surface pollution on target detection,and has strong real-time performance.(2)Aiming at the low sorting accuracy caused by motor speed fluctuation,a multi-target tracking speed measurement algorithm is proposed.The algorithm utilizes the fusion features such as motion law,area,gray level,and corner points to realize multi-target tracking on the belt quickly and accurately,and implement multi-target speed measurement on the basis of tracking.After testing,the tracking accuracy of this algorithm is up to 99%,the speed measurement error is less than 10%,and the efficiency of speed measurement is so high that meets the real-time requirements.(3)A QT-based air jet sorting system is proposed to solve the problem of low color sorting accuracy when sorting masonry and concrete.The system integrates the target detection algorithm and multi-target tracking speed measurement algorithm under pollution conditions,accurately detects the target on-board and calculates the exact time of each target arrives at the air injection valve,reducing the air injection time,improving the air injection accuracy and the color sorter work efficiency.Through repeated tests,we find this system has strong robustness and color sorting accuracy is up to 95%,which meets the basic requirements of accurate sorting of construction waste bricks and concrete.The air jet sorting technology developed in this paper belongs to the common technology of waste sorting and ore sorting.It can be applied to brick-concrete sorting in the field of construction waste.Through this technology,the defective products of ore can be eliminated when comes to ore sorting,and it’s application range is relatively wide.