Schlüsselwort: visuelle Größenkodierung

Due to the currently prevailing worldwide strong desire for innovation and competitiveness, product manufacturers are required to launch technical innovations in short cycles. The computerized society and the global networking, which allow permanent communication and create the desire for a continuous exchange of information, also encourage this rapid development. In the field of communication technology, this constantly progressing introduction and establishment of technical innovations through improved technologies is extremely striking. The eye-gaze interaction, on which this thesis is based, is one of these new communication and interaction technologies. It derives from the niche area of physically disabled people and tries to become more significant as a technical innovation in every day and traditional human-product interactions thanks to the growing computer power. Unlike conventional selection of an object by manual operation of controls or direct touching, in case of eye-gaze interaction it occurs through eye movements, thus by gaze. As described in the literature and in relevant standards, in both general interaction and eye-gaze interaction between the operant and the technical product the user-centered design should have priority at best. Hence, the different user factors, which are structured and detained in standards and policies through defined principles, guidelines, design requirements and design recommendations, should be involved in the design of a new product or in a new interaction technique.The main guideline in this thesis, which is put up for discussion and characterizes the research topic, is the visual feedback of status change. Own preliminary studies as well as other research results prove that the visual feedback of status change requires, particularly for the eye-gaze interaction, the utmost care. For instance, in human-computer-interaction (HCIA) the replacing of the traditional mouse operation by eye-gaze operation has shown that a continuous visual feedback via cursor is disruptive for the eye-gaze based interaction and has an adverse effect on the operation. It is therefore important for the eye-gaze interaction to realize a user-friendly operation with other discreet visual feedback types, which clearly transmit the required status change. The visual feedback of status change by size coding, on which the focus of this thesis lies, is of enormous significance in the ergonomic and user-centered design of an eye-gaze based user interface. Compared to others, size coding is the easiest and the least design interfering coding type. It is advantageous not only for rapid and precise identification of the required status change, but – with the related increasing interaction area especially for eye-gaze interaction – also decisive for an increasing operating accuracy. A research work and analysis of standards and policies about the exact definition of visual feedback as well as an additional evaluation of an analytical expert rating show that size coding is inadequate and only partially quantified. Until this day, there is no literature fully quantifying size coding. For this reason, the aim is to achieve full quantification of size coding by a basic-oriented research study and to give general recommendations for the ideal design of visual feedback by size coding. For the visual size coding increases the security of use by improving the distinctiveness and reduces the operating time. The basic-oriented research study is based on the visual difference detection of geometrical objects with standardized graded object sizes. Thereby the difference thresholds are recorded at appropriate scaling direction (decrease / increase), presentation variant (parallel / serial), varying object size (10 mm / 15 mm / 25 mm), achromatic form-background -contrast (white / black) and varying presentation distance (600 mm / 1000 mm / 1400 mm) of individual object gestalts (circle / square / triangle / pentagon) under laboratory conditions with voluntary test persons. The statistical analysis and interpretation of the results show that the difference threshold is independent of the scaling direction. However, the threshold clearly depends on the presentation variant. At parallel presentation, the difference threshold is greater than the threshold at serial presentation. In addition, the difference threshold depends on the object size, which means the greater the initial object, the smaller the applicable size coding factor to detect a significant size difference. The achromatic form-background-contrast doesn’t affect the difference threshold anymore. After the summary of the difference thresholds to arithmetical averaged values for the purpose of the argued hypotheses, the statement of varying presentation distance is, that the greater the distance between the viewer and the object, the greater the enlargement or reduction factor for the size coding. At last for the different object gestalts, different size coding factors arise, too. Summed up, depending on the object and initial object size, different size coding factors from 1,148 to 1,191 apply for the parallel presentation variant and from 1,067 to 1,098 for the serial presentation variant. This means that an expansion or reduction of 14,8 % up to 19,1 % should take place for visual feedback of status change (parallel) and an expansion or reduction of 6,7 % up to 9,8 % for visual position/pointing or selection feedback (serial). Thus, recommendations are given on the one hand in terms of quantified factors, on the other hand in terms of quantified equations. These are also illustrated in charts which can be used for extraction and calculation of the size coding factors for the respective operating scenario. The obtained result of different size coding factors for each object gestalt has induced to answer the question of what caused this effect. It seems reasonable to suppose that during the detection of a size difference between two object gestalts depending on the object gestalt itself, a different amount of information and conspicuousness are transmitted, which are relevant to the visual difference detection and at last to the size coding factor. To approve this assumption, an evaluation method of statistical-syntactic information and corresponding conspicuousness measure of the used object gestalts is also presented in this thesis. The statement is made, that an object gestalt with its different geometrical elements and elements of order therefore has its own conspicuousness. In relation to the presented quantification of the size coding, the conspicuousness plays an important role for the difference detection. Compared to the other three object gestalts, with 12,8 % the object circle has the lowest conspicuousness, whereof a larger size coding factor results. This means that the object gestalt has an influence on the difference detection and furthermore that the more statistically-syntactic conspicuousness an object gestalt has, the faster the difference detection occurs and the smaller the size coding factor is.