The Development Of Aiming Device Based On Image Processing

In this paper, a design scheme of aiming device is presented. The design is to use the display screen with Six LEDs uniformly distributed, computing the intersection of the built-in camera’s optical axis and the screen. The design of the aiming device can meet the requirements of real-time performance.And the accuracy of the calculation is verified by experimentsSix LEDsare located at the four corners of the screen and the center of the upper and lower edges. The aiming device calculating the position of the virtual nodes needs to get 4 LEDsformed a rectangle on the screen. Aiming device and LEDs are controlled by synchronous signal, and theLEDs located in different positions are alternately illuminated in the odd and even fields. Combined with frame difference we can get a relatively clean image.We use the perceptron model andAdaBoost algorithm in machine learning to judge the LED spots in the image. Firstly, we do the process of image filtering, threshold segmentation and edge detection,and then extract the image sample. Extracting some features that can distinguish the right LED spot edge as input, to build a forecasting model. After obtaining the edge region of LED spot, we take the edge of the center of gravity as the coordinates of the LED spot in the image.After calibrating the camera, we get the camera internal references and the camera distortion factors. Then we correct the LED spots’ coordinates with the distortion factors. We choose four LED spots that can form a rectangle on the screen and imaged in the camera, to calculate the aspect ratio of the rectangle. And then get the perspective transformation matrix to calculate the position of the virtual node with a normalized result.