Research On Optimization System Of Solid Wood Based On Image Recognition

The per capita amount of forest resources in China is very scarce, and wood utilization rate is low. In order to improve the quality and quantity of solid wood material, automated and intelligent optimization equipment for wood lumber should be put into use. However, because of the limitation of technical and economic factors, most wood processing enterprises still use artificial marking method to identify surface defects on solid wood and to determine the cutting programmer, which will lead to many problems. The automation level of wood optimization processing machinery is low, the results of defect recognition are not accurate, and the selection of sawing scheme is unreasonable. In view of solving above problems, this paper presents a study on the optimization system of solid wood plate based on image recognition, and the following main research work has been completed.Firstly, according to consulting relevant references, and combining with the actual collected solid wood plate surface defect images, several common solid wood defect features and effect on wood quality were analyzed; wood plate optimization processing system was elaborated which included several working procedures such as the feed, image acquisition, image processing, optimization calculation, saw cutting, classification of the material. Then the principle and function of the important processes were introduced. The CCD camera and image acquisition card had been selected to build an image acquisition system of solid wood, acquiring two-dimensional color image of wood plate surface. The emphasis on this paper was image processing of wood plate surface defects and optimization algorithm.Secondly, in image processing, the edge detections of two typical images were conducted by using the Sobel operator, Roberts operator, Prewitt operator, Canny operator and LOG operator respectively. The different results were analyzed and compared. Then, image segmentation was realized by using the gray threshold segmentation, histogram threshold, maximum entropy threshold segmentation method respectively. The maximum entropy threshold segmentation method was selected through comprehensive analysis. But the segmentation result was not satisfactory. At last, the basic operations of morphological processing such as opening and closing, dilation and erosion were used to give the segmentation image filling, morphological filtering, edge extraction and mask operation,and acquire closed gray image target. After image processing of the wood plate binary image for defect feature extraction, solid wood surface defect data and information data were obtained including the length and width, the number of defects, defect location and defect dimension information of the wood plate.Thirdly, in order to make the optimal cutting plan, genetic algorithm was used to optimize the calculation. this paper conducted a detailed exposition on principles and the specific steps of the genetic algorithm including expressing problems, generating the initial population, selection, crossover and mutation. The software for optimizing solid wood plate had been developed based on Visual Basic programming tools, which can meet different using needs of preferred mode like material yield, defect sawing, length selection, value optimization and other functions. According to the principle of optimization, combined with the specific examples, the processing scheme of different solid wood plates was given, and the validity of the calculation method was verified.