APN02V10-0892
Introduction
Fuzzy Inference
Objective
Definition of Input/Output Variables
Fuzzy Rules
Source Code of FIU
Input/Output Response
Comments
Cameras with automatic focusing systems usually measure the distance to the center of a finder's view. This method, however, is inaccurate when the object of interest is not at the center of the view (Figure 1). Measuring more than one distance is an approach that may solve this problem. The following example shows the application of fuzzy inference as a means of automatically determining correct focus distance.
Figure 1 Center Based Focusing
Determine the object distance using three distance measures for an automatic camera focusing system.
Definition of Input/Out Variables
Inputs to the FIU (Fuzzy Inference Unit) are three distance measures at left, center and right points in the finder view. Outputs are the plausibility values associated with these three points (Figure 2). The point with the highest plausibility is deemed to be the object of interest. Its distance is then forwarded to the automatic focusing system.
Figure 2a Three Distance Measures
Figure 2b Fuzzy Inference Unit
Each input variable, representing distance, has three labels: Near, Medium, and Far. Each output variable, representing plausibility, has four labels: Low, Medium, High, and VeryHigh. Membership functions corresponding to these labels are shown in Figures 3a and 3b.
Figure 3a Labels and Membership Functions of Input Distance
Figure 3b Labels and Membership Functions of Output Plausibility
The guiding principle for establishing rules of this automatic focusing system is that the likelihood of an object being at medium distance (typically 10 meters) is high, and becomes very low as distance increases (say, more than 40 meters).
To Source Code of Fuzzy Inference Unit
Now let us compile the FIU source code given above and use the FIDE analyzer to see how this unit works. Figures 4a and 4b provides two input/output response surfaces of the FIU. From Figure 4a, we see that Plausibility_of_Center becomes high when the distance at the center is around 10 meters, a distance we defined to be Medium in the definition of input variables. It becomes lower when the distance increases, especially when the distance on the left is Medium. Figure 4b shows the Plausibility_of_Left is high when the distance on the left is around 10 meters, except in the case when the distance at the center is also around 10 meters. In this case, when the distance at the center is about the same as that on the left, we choose center as the desired object. The Plausibility_of_Right is similar to the Plausibility_of_Left. The three outputs of the FIU are compared to identify the point with highest plausibility. The distance at this point is the focus distance.
By adjusting the membership functions of the distance labels, we can achieve different response surfaces for different purposes.
Figure 4a I/O Response : Plausibility_of_Center
Figure 4b I/O Response : Plausibility_of_Left
Remember that this example is provided only for easy-to-use compact cameras targeted for the mass market. For professional photographers it may be inappropriate to provide strictly automatic camera focusing using the three distance measures method. However, if suitable manual overrides were available, it would still be useful as an option in some situations (e.g. when speed is important). Besides automatic focusing(AF), fuzzy logic can be used in automatic exposure(AE) and automatic zooming(AZ). For AE and AZ, the input/output variables and rules of FIU will be different from those shown above for AF, but the design process is very similar.
For Further Information Please Contact:
FuzzyNet http://www.aptronix.com/fuzzynet
Email: fuzzynet@aptronix.com
Weijing Zhang,
Applications Engineer.
Copyright © 1992 by Aptronix
Inc.
Revised: October 21, 1996.