The Research Of Raindrops Spectrum Image Processing System

In nature, the composition of the substance of the tiny unit called the particles. Raindrops particles is the micro forms of substance in rainfall process, that can manifest on the outside the corresponding substance of the physical properties, the chemical properties and statistics rule, etc. Through particle properties can infer the nature of the features and the development trends, according to this characteristic of raindrops spectrum particle image characteristics detect the main parameters in rainfall process. Combined with the machine vision technology and image processing technology, realizing the raindrops features of extraction and statistics, replace the artificial detection, improve the detection of quality and accuracy.This paper design raindrops spectral data detection system, transmit the collected raindrops spectrum image data by the scanner to the computer. Using the computer image processing system analyze and process to the collect images, extract the rainfall containing of raindrops particles, statistics the detection results.In the early image processing, in order to improve the image processing speed, adopting the complex cascade transform method to obtain the effective area of raindrops spectrum sampling image. Interpolation algorithm modify the image distortion problem that is generated the composite cascade transform; The method combining with image enhancement and image filtering, improve the clarity of the target and background image and remove the image noise. Using the method of morphological reconstruction and area filling to solve the phenomenon that is raindrops spectrum particle image of particles hollow and the irregular grain edge, remove particle hollow problem, and realize to processing raindrops spectrum image edge and filling the internal particles, improve the quality of particle edge. Using the Otsu method that is self-adaptive threshold segmentation method achieve the goal particles identification and segmentation, that was evaluated the segmentation effect by the threshold evaluation function.Well-prepared image still exist the phenomenon of the unclear boundary area and the incomplete particle, the paper puts forward image reconstruction algorithm by wavelet analysis, reconstruct waiting raindrops spectrum image between the processed particle image and the original raindrops spectrum image, reduce the noise in the image processing process, then processing the reconstructed image again. On the one hand, this method can calibrate the part of the over-segmentation phenomenon and the noise in the image processing process, on the other hand, raindrops particle image edges became more pliable. In order to improve the precision and accuracy of test, this article from the Angle of edge of raindrops particle testing, firstly, according to the phenomenon, this paper firstly the nature of grain kind round rain, rain particles into a single raindrops particles and adhesion of raindrops particles two kinds. Single raindrops granule with image processing can achieve the processing requirements, and the adhesion particles mainly adopt the Canny of raindrops particle edge detection algorithm to edge detection. The improved Hough transform to detect round parameters, to extract particle radius and the raindrops circular circle. Using statistical theory that the method of undetermined coefficients of raindrops particles data curve fitting, through the least squares solution regression coefficients of numerical analysis that obtain experimental data, thus realize raindrops splash of particle size statistics. Finally, the paper proposes VC and MATLAB mixing programming method,combine MATLAB powerful operation function with VC simple visualization of user interface design function. MATLAB of the rich image processing function library, and VC of the strong function of visual interface, through the interface program that will combine intelligently two software together, achieve raindrops spectrum of man-machine operation interface, improve the simplicity of raindrops spectrum image processing system and strengthen the intelligence of system.Experimental results show that,using machine vision processing raindrop images, it be able to quickly and accurately achieve the raindrop particles segmentation, statistical, and this method has high efficient, easy automated features, simultaneously, there are good practical, so this method provides a new solution for the raindrop of data detection.