Research On Material Data Acquisition And Modeling Method Based On BTF

With the continuous development of society,photorealistic rendering plays an increasingly important role in the field of visual technology.Accurately reproducing the reflective properties of object materials is crucial in achieving photorealistic rendering.For complex materials such as rough surfaces,the Bidirectional Texture Function(BTF)model is widely used to describe the optical characteristics of materials under different lighting conditions.However,the existing BTF material acquisition devices suffer from drawbacks such as low efficiency,poor accuracy,and high cost.Therefore,research on efficient acquisition and modeling of BTF material data is of great significance.This thesis focuses on the study of efficient acquisition and modeling methods of BTF material data.Firstly,a semi-spherical BTF material data acquisition device is designed and constructed using a multi-camera multi-illumination approach.The device features a simple and compact structure,with four industrial cameras and twenty-four evenly distributed LED point light sources on the hemisphere.With this device,we can capture image data of object materials under different lighting and camera positions.The design of the acquisition device ensures data accuracy,and the use of parallel camera capture improves acquisition efficiency.Secondly,a BRDF fitting method based on shadow processing of BTF measurement image data is proposed for material data modeling.This method divides the captured BTF image data into geometric information and the BRDF model.First of all,the geometric information of the material,including the normal map and height map,is computed using photometric stereo.Then,the shadow maps for all captured images are calculated based on the geometric information.Subsequently,existing BRDF fitting models are employed to perform lighting modeling for each pixel.Finally,the material is rendered based on the geometric information and lighting modeling,demonstrating the visual effects of the material.Compared to the method using original BTF data for fitting,using shadowremoved data for fitting and rendering produces brighter and more realistic results.The research presented in this thesis is of significant importance for the efficient acquisition of BTF material data and the achievement of photorealistic rendering.With the self-designed acquisition device,it is possible to capture image data of object materials efficiently and accurately.By utilizing the BRDF fitting method based on shadow processing,better rendering results can be achieved.The findings of this research contribute to making computer-generated images more realistic and closer to the real world.