Research On Optical Adaptive Scene Transformation And Display

In recent years,with the increasing demand of consumers for visual experience,optical adaptive scene transformation and display technology has been applied to a variety of business occasions.Optical adaptive scene transformation and display technology refers to:For the observer in front of the scene,the image signal output by the display device correlates with the observer's position.Thus,we can match the content of the original scene and the restored scene to achieve the visual effect of scene fusion.At the same time,the display parameters of the display device are related to the characteristics of the environment light source,in order to restore the color performance as real as possible.This paper constructs the whole process of scene transformation and display,proposes the display area calculation method based on depth camera and the display parameter control method based on ambient light detection.Then it builds the experimental system,and verifies the effectiveness of this system through experiments.On the basis of the display area calculation method based on depth camera,the system can adjust the display area according to the different position of the observer to ensure the accuracy of fusion.In this method,this paper first analyzes the traditional camera imaging model,and finds that two points whose spatial coordinates are proportional to each other will be mapped to the same pixel during the imaging process.According to the above inference,this paper constructs the coordinate conversion model including three coordinate systems,focusing on the relative position and corresponding relationship of position monitoring camera coordinate system,image capture camera coordinate system and observer coordinate system.The coordinate conversion model needs observer's 3D spatial coordinates as input.So,this paper uses depth camera and MTCNN model to extract position and depth of human eyes,then calculates coordinates of observer's eyes relative to the system.Finally,according to the camera internal parameter matrix obtained in advance,the position deviation of each hardware device in the system and the human eyes' coordinates obtained in real time,the coordinate conversion model is substituted to calculate the reasonable display region.On the basis of the display parameter control method based on ambient light detection,the system can adjust display parameters according to the characteristics of the light source to ensure that the brightness and color of the restored scene are close to the original scene.In this method,this paper first estimates brightness and color temperature of the light source according to RGBC color sensor data,combined with empirical formula and correction formula.Given the brightness and color temperature of the light source,in order to restore the real brightness and color performance of the scene as much as possible,the state control of the display device is divided into three steps.In the first step,a luminance meter is used to calibrate the parameter mapping table of the current display device offline.The second step is to select the closest display parameter in the parameter mapping table according to the current ambient brightness and color temperature.The third step is to call the MCAPI function interface to automatically adjust the device parameters.Finally,this paper builds the experimental system,and evaluates its effectiveness from five aspects:time efficiency,scene fusion visual effect,matching error,brightness deviation and color deviation,contrast between restored scene and original scene.The experimental results show that the adaptive scene transformation and display system designed in this paper can accurately restore scene content in different occasions,and ensure the consistency of brightness and color simultaneously.One transformation loop only takes 283ms,and the display content dislocation is about 1%.In CIE1976L*a*b*uniform color space,the brightness deviation and color deviation are 3.8 and 12.4 respectively.And the contrast between the restored scene and the original scene is 15.8%.