This paper presents a quantitative evaluation of two area-based stereo matching approaches aimed at real-time applications. The former approach follows a standard scheme based on a double matching phase (DMP) and aimed at rejecting unreliable disparities by enforcing the left-right consistency constraint. This approach is representative of several implementations known in literature. The latter approach, proposed recently, relies on a single matching phase (SMF) and rejects unreliable matches by detecting violations of the uniqueness constraints. We review first the two approaches and analyse their differences. Successively, we provide quantitative results, using standard stereo pairs with ground truth, aimed at evaluating and comparing the two approaches in terms of matching reliability and speed.

A MININET "station" acts as a user concentrator into the MININET word-switched network where each word (of up to 16 bits) entering a network "port" is mapped into a single packet for immediate transmission through the network. The stations maintain highly transparent "virtual connections" between user devices. Among the important requirements for MININET are that, under saturation conditions, the communication resources of the network be shared fairly between the user devices and that the network possesses very short end-to-end propagation delays. The design of a station meeting these requirements, capable of processing over one million packets per second, is described. Its architecture is based on special-purpose processors which are functionally distributed. The biggest design challenge was to minimize the processing delay in the node while avoiding potential blockages and remaining fair. The final solution has been to use a master arbiter, which looks at the readiness of the ports and channels in conjunction with connection, routing, and flow control information. The station management entities reside in a general-purpose microcomputer. These monitor and modify the operation of the high-speed processors, establish virtual connections, and support the routing algorithm.

This paper describes a control protocol for high-speed transfers between transceivers in a data communication system. One particularly suitable application of current topical interest would be to control lines linking nodes in a local area packet-switching system. The protocol applies for fixed-length data packets which must ultimately be received at their destination in sequence. With these specifications, a set of rules which is easy to implement in practice is established. The result is a packet interlock method of numbering packet transmissions and jumping back to retransmit a series of packets under error conditions. Problems associated with multiple errors, dummy packets, initial synchronization and resynchronization are investigated and found to be readily overcome by the proposed protocol.