Measurement Of Fuel Spray Droplets Based On Digital Image Processing

The fuel spray is very important to the power performance, fuel consumption, exhaust emission of diesel. Because the size and the quantity of drops is the important index to appraise the spray quality and the spray characteristic, the measurement of the fuel spray drops has the important meaning. Currently, the measure method of the spray drops is varied, because of the cost cheap, and be easy to the operation, the digital camera method has become newly arisen means of measuring the spray drops. This thesis is on the foundation of obtaining the image through the digital camera, adopts the digital image processing and analysis methods to obtain the size and the quantity of the spray drops, is a brave trial and investigate about the study of the spray drops measuring.This thesis study the complete course of the image processing and analysis based on the characteristic of the fuel spray drops-image. The main algorithm is as follows, subtracting background method has corrected the uneven brightness; the pixel operation has overcome the characteristic of the low contrast ratio of the drops-image. Comparing with different threshold segmentation methods and combining theory analysis and experimental result, we has adopted maximum variance method to effectively segment drops-image; the morphology method has resolved the problem of drops adhesion in the drops-image; the recursive searching algorithm has achieved measurement of the drops in the spray drops-image.Meanwhile, this thesis is on the basis of Visual C ++ strong developing platforms and MATLAB abundant toolboxes, develop fuel spray drops measuring system; it is DLL method that realize the mixed programming between Visual C++ and MATLAB. The system owns seven major functions module, namely: image storage, brightness adjustment, image enhancement, image transformation, image segmentation, morphology operation, drops measurement.