Study On Noise Analysis And Preprocessing Of Structured Light Three-dimensional Measurement Image

Structured light 3D measurement is widely used in 3D measurement fields due to its high measurement accuracy,good real-time performance and non-contact.The extraction accuracy of structured light stripe center is one of the main influencing factors of system measurement accuracy.However,various noises in the measurement process will reduce the extraction accuracy of the stripe center,there will be some data with large error in the final measurement result,which reduces the measurement precision of the system.Therefore,there is an urgent need for a scheme to improve the image quality of structured light stripe and an evaluation method for the quality of the extraction results of structured light stripe.This paper mainly studies the structured light stripe image,analyzes the noise of the structured light system.On the basis of the analysis,an image processing method for improving structured light stripe and an image quality evaluation method are provided to improve the measurement accuracy of the system.The specific work is as follows:Firstly,the basic principle of three-dimensional measurement of structured light is given.The noise is systematically analyzed from the causes of structured light image noise,and detailed analysis is carried out from projection,measured object,shooting,environment and so on.Finally,the noise is classified and summarized.Aiming at the degradation of stripe gray distribution image,an image processing scheme is proposed to improve the accuracy of stripe center extraction.An witching median filtering scheme and a Gaussian filtering method are used to filter out noise while ensuring the image details as much as possible.This paper analyzes the common methods of stripe center extraction,and a stripe center extraction method is proposed,which is based on the principle of equal energy values on both sides of the stripe center.The advantage of this method is that it is less affected by the asymmetric distribution of the image.Due to the interference of the structured light stripe image noise,the characteristic gray distribution of the stripe pattern degenerates from the original symmetrical distribution to the asymmetric distribution,which seriously affects the accuracy of the stripe center extraction.Therefore,there may be data with large error in the measurement data.This paper proposes a method for evaluating the quality of structured light stripe image.The method proposes the stripe center location error and the skewness coefficient based on the third-order central moment,and evaluates the quality of stripe gray distribution by using the skewness coefficient of the stripe gray distribution.By setting the threshold of the skewness coefficient,the data with larger error can be identified and directly eliminated,and the setting method of the threshold of the skewness coefficient is given.Simulation results verify the feasibility of this method.Finally,the feasibility and effectiveness of the above method are verified by simulation and experiment.