Quality Grading Evaluation Of Cut Roses Based On Computer Vision

With more and more people's pursuit of visual enjoyment and warm living environment,people's demand for flowers is growing,and the flower industry has gradually become an emerging industry in China.As one of the most popular flowers,cut roses is increasing year by year,and the quality problem has become more and more important.It has become the most concerned and concerned issue for flower farmers,the production enterprises and the auction market of cut roses.Based on the quality classification standard of cut rose flowers,this thesis takes the quality of f cut roses as the research object,applies computer vision method to extract various indicators,and then comprehensively evaluates each index based on factor analysis method,and Finally,a quality grading evaluation model of cut roses based on Support vector machine is constructed to predict the grade of cut roses.(1)Detecting indicators related to cut roses.In this thesis,the image collection system of cut roses was first built,and then the image processing of cut roses was collected,including image denoising,image segmentation,grayscale processing,binarization processing and mathematical morphology processing.a boundary-clear binary image,and the edge of the image is obtained by edge detection;then the minimum circumscribed rectangle and the minimum circumscribed circle of the image are obtained,and finally 17 characteristic index values are obtained;(2)Identify evaluation indicators and expert data.Firstly,the factor analysis method is used to analyze the 17 indicators cut roses,and the common factors are extracted.Then,the comprehensive scores of cut roses samples are calculated by the proposed common factors,and the samples are classified by comprehensive scores to form expert samples.(3)Constructing a quality evaluation model for cut roses.In this study,a quality grading evaluation model is constructed based on the support vector machine classification algorithm,and the parameters are optimized by the grid search algorithm.The accuracy of the final model reached 87%,which achieved a good prediction effect.At the same time,compared with other classification algorithms,the comparison results show that the model based on support vector machine has better effect,indicating that the theory and model used are It is feasible and practical to establish the quality evaluation model of cut roses.