Humanoid-Eye Imaging System Model with Ability of Resolving Power Computing - Advances in Image and Graphics Technologies - page 99

We define these two items as judgment of wave optics and judgment of receptors,
which indicates resolution limit thresholds by the two criteria. Note that these judg-
ments are determined by inherent feature parameters of HIS and by these judgments,
resolving power of different applications of HIS can be assessed. For implementations
in a specific imaging system, the comparison of the two judgments should be made
first. Provided all outside conditions are the same, the set of formulas for ARP and
MDD of which the judgment is larger is chosen as resolving power calculation for the
system and the smaller one's relevant criterion will be automatically satisfied.

4 Experiments

To prove the formula of APR and MDD for HIS, an experiment is carried out with a
digital camera (Canon PowerShot A710 IS) as the practical implementation of HIS
and two LEDs as object points.
Relied on parameters of the camera, we calculate the judgments of the system to

=× . The resolution limit threshold by theory of
receptors far exceeds that by wave optics. Therefore the actual resolving power calcu-
lations for ARP and MDD are Eq. (8) and Eq. (9).

J = 4.538 × 10

− 5

R J −

4.05 10

4

get and

WO

4.1 Procedure and Results

In the experiment, the effects of factors d 0 , Ψ and ʳ are proved individually for the
static situation. For dynamic situation, the effect of d 0 is tested and effects of Ψ and ʳ
are evaluated together, with different speed Vc settings for each sampling data
groups. The actual approach is to calculate MDD by given parameters, then set the
distance of the two LEDs to be the predicted MDD and sample images to see whether
the prediction will match the results. By setting an area of 3×3 pixels as the equivalent
receptor, two points are just distinguishable when the areas of 3×3 pixels where
LEDs' images are located are right separated by an equivalent receptor. Some adjust-
ment is made on original images for a better localization of LEDs. Some of the results
are showed in tables 1 to 4, including original images (upper ones in each image set)
and adjusted images (lower ones in each image set).

Table 1. Predicted static MDD and results for the effect of d 0

1500 3300 6300

d / mm

O

3.33 7.33 14.00

 s / mm

image