Real-time colour recognition in symbolic programming for machine vision systems

Abstract

Abstract:It is impossible to collect more than a tiny proportion of all of the possible examples of a given hue, to form a training set for a machine that learns to discriminate colours. In view of this, it is argued that colour generalisation is essential. Three mechanisms for learning colours as defined by a human being are described. One of these is based upon an idea developed by AP Plummer and is implemented in a commercial device known as the Intelligent Camera. This implementation can learn the characteristics of coloured scenes presented to it, and can segment a video image in real-time. This paper presents four procedures which allow the range of colours learned by such a system to be broadened, so that recognition is made more reliable and less prone to generating noisy images, which are difficult to analyse. Three of the procedures can be used to improve colour discrimination, while a fourth procedure is used when a single and general colour concept has to be learned. Several experiments were devised to demonstrate the effectiveness of colour generalisation. These have shown that it is indeed possible to achieve reliable colour discrimination / recognition for such tasks as inspecting packaging and fruit. A practical system based upon the Intelligent Camera and controlled by software written in Prolog has been developed by the authors and is 2 being used in a study of methods for declarative programming of machine vision systems for industrial applications.Key words: Machine Vision, Systems Engineering, Teaching by showing, Realtime colour recognition, Prolog. Introduction Colour Recognition in a Symbolic Programming EnvironmentConsider the task of designing a machine to inspect printed cardboard and plastic cartons, such as those used for food products, household goods, toiletries, etc. These are often printed in several colours. For example, the container of a certain brand of margarine has a wavy yellow streak across its top, in addition to the product name, which is printed in red. Other features on the top are printed in blue (large blob-like regions) and black (small lettering). (See Figure 1.) Inspecting such a container could be achieved by first isolating the different colours and then applying conventional (i.e. To illustrate how colour recognition may be used in symbolic programming of a vision system, we shall explain how we can identify bananas, which is a rather simpler task than inspecting margarine tubs, which we shall discuss in detail later.The following is a description of a banana, expressed in formalised English.X is a banana if X is yellow and the length of X is U and % U is the length in mm 60 ≤ U ≤ 800 and the width of X is V and % V is the width in mm 10 ≤ V ≤ 50.This recognition rule for bananas 1 can be translated into Prolog in the following way.(Prolog has been used for several years as a medium for programming vision systems. (Batchelor 1991;Batchelor and Whelan 1993; and will be discussed again in this article.) object(X, banana) :-colour(X,yellow), length(X,U), 60 ≤ U, ...

<title>High-resolution image digitizing through 12x3-bit RGB-filtered CCD camera</title>

“…The color recognition171 [8], sampling and processing9 are essential to the color image capturing. The color signal from discrete RGB frames have to be integrated to form actual color.…”