Research On Related Techniques Of Vision Inspection Based On Line Structured Light

Structured light vision inspection technology was put forward to solve the image matching problem in three-dimension visual. Recently this technology has attracted a great deal of attention in industrial environment for its some good characters, such as its high precision, easy to abstract the information from laser stripe image and strong real-time ability and so on. This technique is too important to be replaced by others in the fields of 3-D auto-measurement and object recognition.According to the development and practice of Modern Shipbuilding Huge Size 3-D Numeral Measurement System, this dissertation has carried out research deeply on the such problems as image processing technology, laser stripe edge detection, measurement system model establishment and the factors that affect the accuracy of the model. These problems directly affect the measurement accuracy of the measurement system.One of those key tasks of detection is the processing and calculating of the structure light image. In Computer Vision Systems, visual information processing technology depends on image processing method which can obviously improve the quality of image after processing. Not only the visual effects of image can be improved, but also provide convenience to computer image analysis, processing and recognition. Based on the analysis of the characteristics of the image and the system, median filter was used for image preprocessing. This method can effectively inhibit pulse interference and punctate noise and can better maintain the image edges, stable operation and high processing rate. Edge detection is very import for center coordinates of the stripe, impacting to the accuracy of the center position directly.This dissertation has discussed the principle of vision sensor with structured light, the structure and model of structured light vision system, went deep into analysis about the influence to accuracy from the structure. The major factors that influence the accuracy are speckle, scatter error and edge of laser. This dissertation has carried out research deeply on these factors and discussed the corresponding solutions.There has been growing interest of wavelet-based denoising schemes and edge detection techniques recently. Such popularity is mainly due to that wavelet provides an appropriate basis for separating noise signal from image signal and makes noise and edge show different attributes. A denoising algorithm based on wavelet transform and median filtering is proposed, which is very effective to reduce noise mixed with salt and pepper and gauss noise. This dissertation has also carried out research on an edge detection theory based on scale multiplication in wavelet domain, this algorithm applies well for image edge detection, just as for laser light image which has two or more edges in neighboring field.