Research On Image Fusion Method For Soybean Canopy Information Acquisition

Crop growth information monitoring,using canopy information to reflect is still the main means,thanks to the rapid development of UAV remote sensing technology and computer technology,the use of UAVs with multispectral remote sensing cameras to obtain crop canopy information has become one of the indispensable technical means of crop growth process growth monitoring,but remote sensing images are affected by factors such as flight height and imaging resolution,resulting in poor clarity of the synthesis results and However,remote sensing images are affected by factors such as flight height and imaging resolution,resulting in poor clarity and low amount of effective information,which directly affect the judgment of crop growth information and monitoring efficiency.The research of canopy image fusion method can realize the accurate extraction of crop growth information parameters and provide theoretical and practical significance for the development of crop growth judgment and precision application technology.In order to improve the quality of UAV low-altitude remote sensing imaging,based on the analysis of Chinese and foreign literature,we systematically summarize the current status of research on multi-source remote sensing image fusion algorithms,draw on existing fusion mechanisms,take soybean canopy as the research object,study the soybean canopy remote sensing image fusion algorithm based on the characteristics of soybean canopy UAV low-altitude remote sensing image imaging,obtain a fused image with high resolution and rich spatial detail information The research focuses on the following aspects.(1)Theoretical analysis of low-altitude remote sensing image fusion of soybean canopyThe article compares and analyses the advantages and disadvantages of the image fusion algorithms based on component replacement and multiscale transformation,selects the IHS change in the component transformation and the NSCT transformation based on multiscale,constructs the structure of the fusion image algorithm,determines its working principle,analyses in detail the transformation process of the intensity components in the fusion process,and corrects the information loss and spectral distortion generated by the component transformation by The NSCT transformation operation is combined with the fusion algorithm to correct the loss of information and spectral distortion,effectively avoiding the loss of spatial details and maintaining the representation of spectral information.(2)High and low frequency sub-band coefficient fusion rulesBy analysing the working principles of non-downsampling pyramidal filtering and nondownsampling directional filtering in the NSCT transformation process,the fusion operation of low and high frequency sub-band coefficients between images is completed by introducing the concept of spatial frequency and regional energy weighting,and the image element with the largest spatial frequency is selected to participate in the sub-band reconstruction,so that most of the energy information is retained and the clarity of the results is kept stable;the regional energy weighting based on the intra-regional pixel The regional energy weighting method,which is based on the premise of intra-regional pixel correlation,can also ensure the clarity of the fusion results,while adequately expressing the local features of the image.(3)Experimental study on the application of fusion resultsBased on the theoretical study of image edge segmentation extraction and the actual experimental operation process,the edge extraction method based on search and the edge extraction method based on zero-crossing are analysed respectively to verify the edge extraction effect of the fusion results,while the vegetation index obtained from the fusion results is experimentally verified by combining the results of the near-earth remote sensing vehicle sensor.The research establishes an IHS transform combined with non-downsampling contour wave transform for UAV low-altitude remote sensing image fusion algorithm,and applies UAV remote sensing directly to soybean canopy information extraction.The fusion results have high spatial information retention and good image resolution,which provides a theoretical basis and technical support for soybean canopy image edge contour information extraction and vegetation index acquisition.