Objects Floyd Nelson A.D Dec 29
Examples of Objects Representation: desk – Object: brown, with four brown lines branching off from it Representation: loud – Object: noise that is loud Representation: name – Object: qualities that describe the name objects are usually pictures, but can involve 1 dimensional inputs and/or multiple inputs, outputs, etc. an object is a reference to a kind of experience
An Object’s Elements when describing a new object these qualities are used – pictures from views of the object – average color – “rotation shift” when describing a current object these qualities are used – rotation – size – stretch – *rectangular trapezoidal
Objects to Look for Most objects do not disappear (except for objects moving on computer screens, out of vision or behind a closer object). – So the system should check for objects that were there previously. Rooms usually have the same items/objects in them. – So the system should check for objects that are usually in the location that it is in. – The system’s location can be found by: knowing some of the objects in the room knowing what objects it saw before it entered the room Objects may appear to turn into other objects. – A pencil looks like a long yellow rectangular shape; after it is rotated, it can look like a small pink circle If an object is mentioned from communication, the system should probably search for that object. Anything that changes (moves)
Identifying Objects Use color difference to generate an outline of the object Find the longest point-to- point of the object Compare with other objects. If radically different from all other objects, make a new object. Use touch information to determine 3D shape There are several different kinds of objects for each word in human languages. – There is more than one shape of computer screens and chairs. – When someone thinks of a dog, they will not always think of the same dog, even if it is the same species of dog (one could be fatter).
Methods of Identifying Objects 1. To find the range of pixels to work with: 2. Find its longest side and its bisector using a line test of varying angles – The minimum (from side) and maximum (from diagonal) are identified from the first test 3. Overlap the two images; the colors in fig. C continually alternates (green, red, green, red, …) blue / (red + green + blue) = % similar
Oddities 1 Counting the number of members of a flock – Count each individual while remembering which ones were counted – Area × Average_Density = number_of_Members
Oddities 2 Disappearing and appearing objects & rapid change of color – iguanas – glow in the dark – light bulbs – night/day – television – computer screens
oddities 3 blur fog rainbow How can the perimeter of the object be determined?