### Table 5.2: Vision Algorithm Performance Summary

1996

Cited by 4

### Table I and Fig. 9(c) illustrates the result of candidate arcs evaluation without GPS inputs. The vision algorithm ranks three top choices including arcs 3, 4, and 5. Figure 9(d) show how a GPS signal is used to identify the final choice. Figure 10 uses two more examples to further illustrate how a GPS signal can be used to improve the quality of the output of the vision algorithm. Note that the starting points of the arcs in all examples are calculated with considering vehicle kinematic constraints and time-delays.

### Table 1 summarizes the status of the project. In general terms, most of the project goals regarding computer vision algorithmic have been achieved, while the end-user part remains to be done in 2006. It is obviously the last part of the project because it depends on the reliability of the tasks developed in the steps 1.1, 1.2 and 1.3.

"... In PAGE 5: ... Table1 : Status of the project goals at August, 2005 2 Tasks and development In this section we detail the progress of the tasks related to each of the project goals, as well... ..."

### Table 2: Performance comparison with stereo algorithms from the Middlebury Stereo Vision page [2] for the Map im- age pair. MAE: Mean Absolute Error, RMS: Root Mean Square error.

"... In PAGE 4: ... We also report the results of some state- of-the-art methods from the same website. Table2 provides a quantitative comparative study using two global error mea- sures: the Root Mean Square error and the Mean Absolute Error. We note that in accordance with [2], we exclude pixels that are in occluded regions when computing the disparity errors.... In PAGE 4: ...arity map (Fig. 3.c), this would not have been necessary for our approach as it yields robust disparity estimates in those points. In order to measure the impact of each constraint, we also indicate in Table2 the error values obtained when one of the constraints is missing. This confirms that it is useful to incorporate multiple constraints.... ..."

Cited by 1

### TABLE I ROOT MEAN SQUARED ERRORS FOR POSITION (IN RADIANS), VELOCITY (IN RADIANS/SEC) AND FEEDBACK COMMAND (IN NEWTON-METERS) FOR THE SARCOS ROBOTIC VISION HEAD. ALGORITHMS EVALUATED INCLUDE RIDGE REGRESSION WITH NONLINEAR GRADIENT DESCENT, OUR BAYESIAN DE-NOISING ALGORITHM, LASSO REGRESSION WITH THE PROJECTION STEP, AND STEPWISE REGRESSION WITH THE PROJECTION STEP. STANDARD DEVIATIONS ARE NEGLIGIBLE AND THUS OMITTED. Position (radians) Velocity (radians/sec) Feedback Command (Newton-meter)

2006

Cited by 2

### Table 2. Performance of the prototype machine vision system using ELG and CMP

"... In PAGE 10: ... For the bad group, . a total of 16 images and 46 images were used respectively Table2 . After the true class of each leaf was determined by manual inspection, the performance of the image processing algorithm was determined using a Bayesian discrimination proce- .... In PAGE 16: ... After several preliminary tests, two features, ELG and CMP, were found to provide the optimal Bayesian classifier for the images in this study. Table2 shows the performance for both image groups with the Bayesian classifier built using these features. Adding more features to the classifier had inconsistent results but in no case did it help reduce the total error rate of the classifier.... ..."

### Table 2: Optical structure characteristics of the MDOF active vision system to make continuous, small optical adjustments required by many algorithms in near real time with excellent preci- sion. Qualitative improvements in lens performances in- crease the advantages of active vision techniques that rely on controlled variations of intrinsic parameters. 2.2 System Architecture The MDOF active vision robot head is controlled by one host computer with a Pentium CPU (166Mhz) and a PCI Matrox Meteor frame-grabber PC board.

1997

Cited by 11

### Table 2: Optical structure characteristics of the MDOF active vision system to make continuous, small optical adjustments required by many algorithms in near real time with excellent preci- sion. Qualitative improvements in lens performances in- crease the advantages of active vision techniques that rely on controlled variations of intrinsic parameters. 2.2 System Architecture The MDOF active vision robot head is controlled by one host computer with a Pentium CPU (166Mhz) and a PCI Matrox Meteor frame-grabber PC board.

1997

Cited by 11

### Table 1: Assumptions in computer vision: Natural constraints (N), physical constraints (P) and synthetic constraints (S).

1997

"... In PAGE 4: ...Table 1: Assumptions in computer vision: Natural constraints (N), physical constraints (P) and synthetic constraints (S). surface smoothness) which are used to convert the ill-posed problem of vision to a well-posed problem (see Table1 (7)). Even well-posed problems may turn out to be computationally in- tractable because of the iterative algorithms used today to solve them (e.... ..."

Cited by 3

### Table 6: A framework for gradient descent image alignment algorithms. Gradient descent image alignment algorithms can be either additive or compositional, and either forwards or inverse. The inverse algorithms are computationally efficient whereas the forwards algorithms are not. The various algorithms can be applied to different sets of warps. Most sets of warps in computer vision form groups and so the forwards additive, the forwards compositional, and the inverse compositional algorithms can be applied to most sets of warps. The inverse additive algorithm can only be applied to a very small class of warps, mostly linear 2D warps. Algorithm For Example Complexity Can be Applied To

2004

"... In PAGE 30: ... In Section 3.4 we validated this equivalence empirically. The four algorithms do differ, however, in two other respects. See Table6 for a summary. Although the computational requirements of the two forwards algorithms are almost identical and the computational requirements of the two inverse algorithms are also almost identical, the two inverse algorithms are far more efficient than the two forwards algorithms.... In PAGE 49: ...pproximations, and the Levenberg-Marquardt approximation. These two choices are orthogonal. For example, one could derive a forwards compositional steepest descent algorithm. The results of the first half are summarized in Table6 . All four algorithms empirically perform equivalently.... ..."

Cited by 144