Study On Detection And Evaluation Of Vegetation Ecological Environment In Open-pit Mining Area Based On Machine Vision

Detection and evaluation of vegetation ecological environment damage in open-pit mining area is an important basis for ecological restoration and treatment of mining area.The vegetation coverage around mining area is an important index to evaluate the health of local ecological environment,and the vegetation coverage is also gradually applied in the ecological evaluation of mine area.Extraction of vegetation information in mining area is the basis of vegetation coverage calculation and calculation and prediction of vegetation coverage is the key to analyze vegetation environmental change.In this paper,based on the method of machine vision to detect and evaluate the ecological vegetation environment in open-pit mining area,the main work is as follows:(1)Proposed a vegetation extraction method based on improved Deep Lab V3+ and attention mechanism.Aiming at the problems of feature loss and blurred edge of target area in traditional vegetation extraction methods in opencast mining areas,densely connected ASPP modules with different cavity rates were constructed to expand the sensitivity field of feature points and improve the coverage of multi-scale features.The adaptive attention mechanism was used to obtain more information features of vegetation images in mining areas,and the Mobilenet V3 with expanded convolution was introduced to reduce the number of network parameters and the amount of calculation,so as to improve the efficiency of target extraction of vegetation areas.The improved vegetation extraction model can effectively improve the segmentation accuracy and effect of vegetation areas.(2)Proposed the NDVI-DFI pixel three-component model to calculate the vegetation coverage of open-pit mining area.Aiming at the problems of slow speed and low accuracy in the calculation process of traditional model,the vegetation extraction results of open-pit mining area were used to optimize the vegetation extraction speed,and the pixel three-component model was selected to estimate the vegetation coverage.By extracting vegetation information,land types were divided,and the open-pit mining area and vegetation covered area were distinguished,so as to improve the accuracy of vegetation coverage estimation.(3)Proposed a prediction model of vegetation coverage based on SSA-LSTM mining area.According to the results of vegetation extraction and coverage estimation in open pit mining area,the prediction model was used for change analysis.Meanwhile,compared with other models,SSA was used to optimize the parameters of LSTM model with the best prediction effect,and the model fit coefficient was the highest.By using the vegetation coverage rate of a long sequence and considering the factors affecting vegetation analysis,the vegetation prediction effect map of the study area is obtained,so as to intuitively show the change of vegetation ecological environment in the open-pit mining area.The experimental results show that the vegetation extraction model established in this paper has a good effect on the speed and accuracy of vegetation extraction detection.At the same time,the vegetation coverage estimation and prediction model is proposed,which can also estimate the results well.Considering the influencing factors of vegetation,the vegetation coverage change from 2022 to 2023 can be calculated and predicted with good effect,which can evaluate and analyze the vegetation ecological environment around the open-pit mining area and provide reference for the restoration and management arrangement of regional ecological vegetation.