外文翻译--三维坐标和颜色信息匹配的3d颜色传感器(编辑修改稿)

外文翻译--三维坐标和颜色信息匹配的3d颜色传感器(编辑修改稿)内容摘要：

techniques are proposed such as Direct Linear Transformation Method, FullScale Nonlinear Optimization Method, Two Stage Method and so on[4, 5]. In most of these methods, plicated camera model always need to be set up and many camera’s intrinsic and outside parameters need to be calculated, which always results in a plex and sometimes instable solution procedure. However, in many applications, the mapping relationship between space points’ coordinates and their pixels’ coordinates in the image is enough and camera’s intrinsic and outside parameters are redundant. Based on this concept, calibration technique with Linear Partition Method and BP Neural Network Method to 3D sensor and color sensor respectively is proposed and the matching between 3D and color information is realized. 2. PRINCIPLE OF CALIBRATION AND INFORMATION MATCHING Linear partition method and its solution procedure 3rd International Symposium on Instrumentation Science and Technology Aug. 18~22, 2020, Xi’an. China 1810 1 1 T The mapping relationship between object’s space coordinates (XW, YW, ZW) and their corresponding pixels coordinates (Xf, Yf) got from image capture process can be formulated in matrix form with homogeneous coordinate as following equation. ⎡ Xw⎤ ⎡ Xw⎤ ⎡ X f ⎤ ⎡ m11 m12 m13 m14 ⎤ ⎢ ⎢ ⎢ ⎢ ρ ⎢ Y ⎢ = ⎢ m m m m ⎢ • ⎢ Yw ⎢ = M • ⎢ Yw ⎢ (1) ⎢ f ⎢ ⎢ 21 22 23 24 ⎢ ⎢ Zw⎢ ⎢ Zw⎢ ⎣⎢ 1 ⎢⎥ ⎢⎣ m31 m32 m33 m34 ⎢⎥ ⎢ ⎢ ⎢ ⎢ ⎣ ⎥ ⎣ ⎥ Where ρ is a scale factor. Apparently, the matrix M contains all of the mapping information and if the number of the calibration points is enough, M can be determined by solving a linear system of equations, which can be created by using calibration points’ 3D coordinates and their corresponding image coordinates. Equation (1) can be expanded as following equation. ⎧ m11 Xw + m12Yw + m13 Zw + m14 − m31 X f Xw − m32 X f Yw − m33 X f Zw = m34 X f ⎨ ⎩ m21 Xw + m22Yw + m23 Zw + m24 − m31Y f Xw − m32Y f Yw − m33Y f Zw = m34Y f (2) Theoretically, parameters from m11 to m34 can be determined by 6 points. However, in practical application, m34 always be treated as one and dozens of calibration points are introduced to reduce the error by solving overdetermined equations. So when the number of the points is N, 2N equations can be obtained and expressed as following and the matrix M can be got from leastsquares procedure. ⎡ Xwi A = ⎢ Ywi Zwi 1 0 Ax = B 0 0 0 − X fi Xwi − X fiYwi − X fi Zwi ⎤ ⎢ (3) ⎣ 0 0 0 0 Xwi Ywi Zwi 1 − Y fi Xwi − Y fiYwi − Y fi Zwi ⎥ X = [m11 m12 m13 m14 m21 m22 m23 m24 m31 m32 m33 ] It doesn’t take into consideration of the distortion of the lens and other nonlinear factors during above discussion and calibration technique simply based on this method will cause much error. So another method is proposed, which divides the whole image into several parts and that also means the data pairs (space points’ coordinates and their corresponding pixels’ coordinates) are divided in to several sets, linear method mentioned above will be applied to each set of data pairs or image region respectively. Several transformation matrixes will be got from this method and when measurement is needed, each of them will be used to certain input data set based on the classification rules of region division. This is the basic concept of Linear Partition Method and measuring error will be reduced significantly by this technique. BP neural work calibration technique BP Neural Network is a oneway transmission and multilayer artificial work. Every layer contains one or more nodes and the output of each layer is only connect with the input of the next layer and have no relationship with the nodes of other layers and itself. A standard work i。