### Table 3 Nominal values of cable attachment points at the Moving platform in the Moving coordinate frame

### Table 4 Real values of the cable attachment points at the Moving Platform in the Moving platform coordinate frame.

in Pathway

"... In PAGE 3: ... 5. SIMULATION PROCEDURE AND RESULTS We simulated the calibration method on a cable robot whose nominal parameter and real parameters are given in Table 1 through Table4 . At present we do not any consistent (real) sensor data to validate our calibration model.... ..."

### Table 1 Comparison of the number of complex arithmetic operations for di!erent moving sizes (M) and frame sizes (N)

1999

"... In PAGE 6: ...505!-(ADD#) is (Eq. (A.8)) N505!-(ADD#) quot;M#(p#1)N. (8) Table1 lists the N505!-(ADD#) and N505!-(MUL#) for di!erent moving sizes and frame sizes. The shaded rows displays the numbers of complex mul- tiplications and additions without implementing the recursive procedure.... In PAGE 7: ... However, it provides better syn- chronization ability with the real data acquisition process. In comparison with the numbers listed in the shaded rows for each given N, the integrated approach does not advantage the computational e$ciency as the moving size M increases (the right side of the bold-faced lines, Table1 ). Considering the example of N quot;512 and M*64, one would rather apply the real-time FFT algorithm without implementing the recursive procedure.... ..."

### Table 13 shows both the percentage of frames that differ between the phone-derived and asynchronous feature transcriptions and the number of boundaries that have shifted. Close to half of all boundaries have moved, resulting in changes to 2.4% of all frame-level feature value labels.

### Table 2: Errors for the pattern in Figure 2 moving of accelerated motion (varying flow). Frame- frame displacements in pixels along the co-ordinate axis are [11111111111222222222 2222223333]and[111111111111111111112222223333]respectively. See text for error definitions and motion parameters.

"... In PAGE 6: ... Conversely, errors increase with the memory span in the case of non-stationary flow. Table2 shows results computed in the same conditions as the previous experiments, but with accelerated motion. We notice that the algorithm is stable in DB,inthe sense that small variations of DB result in small variations of the estimated flows.... In PAGE 8: ...Table2 are reduced approximately by two orders of magnitude. 5.... ..."

### Table 1: Table showing the di erent frame rates obtained for a scene with moving vehicles. We compare our algorithm to others using just Phong shading, or with the complete re-computation of environment maps for each moving object.

2003

Cited by 1

### Table 1: Performance of moving object detection algorithm

2004

"... In PAGE 6: ... The thick rectangles indicate the position of manually-tracked objects, the thin rectangles indicate the output of the tracking algorithm, and the thin lines show the distance between the center of the rectangles. The nal evaluation result is shown in Table1 . Motions is the number of moving objects over the total number of frames.... ..."

Cited by 11

### Table I. Segmentation results comparison. The testing video is ETRI od Bmrre . from frame 101 to frame 221, includina 41 - ~ .I I I- and P-frames. Ground truth is obtained manually. The number of the total blocks of the moving object is 2137.

### Table 1 lists the N505!-(ADD#) and N505!-(MUL#) for di!erent moving sizes and frame sizes. The shaded rows displays the numbers of complex mul- tiplications and additions without implementing the recursive procedure. In that case, the algorithm simply constructs the sr-FFT structure using the real-time implementation method. Note that the real-time FFT algorithm [10] has about the same order of computational complexity as the normal

1999

"... In PAGE 7: ... However, it provides better syn- chronization ability with the real data acquisition process. In comparison with the numbers listed in the shaded rows for each given N, the integrated approach does not advantage the computational e$ciency as the moving size M increases (the right side of the bold-faced lines, Table1 ). Considering the example of N quot;512 and M*64, one would rather apply the real-time FFT algorithm without implementing the recursive procedure.... ..."

### Table 1. Frame-based registration error between supine and prone positions. Patient 06 was excluded because the patient moved during the prone image acquisition.

"... In PAGE 6: ... This patient was therefore excluded from further analysis. The results of the quantitative analysis of the frame-based registration error between supine and prone positions are listed for each patient in Table1 . The maximum error was 2.... ..."