### TABLE I. Tracking Accuracy Comparison Trajectory Validity

### Table 2. Non-Model Based Control for Circular Trajectory Tracking errors during the second iteration ( is optimized, = 0:9 is used)

### Table 3: Subre ector trajectory for gravity-correcting focus tracking

"... In PAGE 14: ... This should essentially eliminate variation of gain with elevation, except for spillover changing as the structure distorts due to gravity. Table3 shows the subre ector actuator trajectory which will implement the optical prescription which was shown in Table 1. The trajectory is presented in two equivalent forms, (1) six displacements and (2) six actuator lengths.... In PAGE 15: ...Figure 5: Gregorian focus tracking actuator trajectory ( Table3... In PAGE 16: ... At E = 90 (zenith), the situation is reversed: the vertical feedarm springs back, so X1 and X2 must extend, while the horizontal feedarm sags under the weight of the vertical feedarm, so Y1, Y2, Y3 must shorten. The reader can verify by inspection that these properties hold in Table3 . The signs of ZS in Table 3, judging by the di erential motions of Y1, Y2 and Y3, also appear to be plausible.... In PAGE 16: ... The reader can verify by inspection that these properties hold in Table 3. The signs of ZS in Table3 , judging by the di erential motions of Y1, Y2 and Y3, also appear to be plausible. 3.... In PAGE 16: ... The direction sense of his XS is ipped relative to the author apos;s intuition. His total travel range is signi cantly di erent from that shown here in Table3 . The most signi cant problem for the author in comparing Enterline apos;s table with the present work is that Enterline presents his results with no explanation of his optical analysis.... ..."

### Table 1. Model Based Approach for Circular Trajectory Tracking errors after convergence ( is optimized. = 0:9; nH = 6; wk ij(0) = 0)

"... In PAGE 9: ... NN controller. The improvement achieved by the NN controller is clearly demonstrated. Trajectory tracking performance is measured by the following sums of tracking error-squares over one training cycle of a trajectory: Ep = P=T Xt=1 k xd ? x k2 (m)2; Ev = P=T Xt=1 k _ xd ? _ x k2 (m=s)2 (52) where Ep is the position error; Ev is the velocity error; P = 4sec is one training cycle elapse time, and T = 5ms is the sampling period. Table1 summarizes the converged performance results of the three proposed schemes. It is clear that performance of NN compensated schemes are much better than that of uncompensated one.... ..."

### Table 3. Non-Model Based Control for Circular Trajectory Tracking errors during the second iteration ( is optimized, = 0:9 is used) Schemes RT CC F CC Uncomp

"... In PAGE 9: ... We also tested proposed algorithms for non-model based case. For non-model based control approaches, the results are listed in Table3 . In this case.... ..."

### TABLE I TRACKING ERROR FOR A 100HZ SINUSOIDAL TRAJECTORY.

### Table 5. Comparison of the potential trajectory number and tracked candidate number on the 86 testing sequences. KF algorithm Proposed PB algorithm

in A Trajectory-Based Ball Tracking Framework with Visual Enrichment for Broadcast Baseball Videos *

### Table 1 summarizes the tracking errors for the trajectories discussed in this work. The best condition refers to the situation in which the first winners for each state of the trajectories are used to retrieve them.

"... In PAGE 6: ...11] P.I. Corke. A robotics toolbox for MATLAB . IEEE Robotics and Automation Magazine, 3(1):24-32, 1996. Table1 . Summary of tracking errors I-G1: 5.... ..."

### Table 1 summarizes the tracking errors for the trajectories discussed in this work. The best condition refers to the situation in which the first winners for each state of the trajectories are used to retrieve them.

"... In PAGE 6: ...11] P.I. Corke. A robotics toolbox for MATLAB . IEEE Robotics and Automation Magazine, 3(1):24-32, 1996. Table1 . Summary of tracking errors I-G1: 5.... ..."