PCL(ID:6215/pcl008)

Pyramid C Language 

for Pyramid C Language

Language for controlling and manipulating pyramids
Related languages
C => PCL   Extension of
PCL => PICL   Extension of
References:
  • Di Gesu, V. "A High Level Language for Pyramidal Architectures" view details
          in Proc. NATO ASI Workshop on Pyramidal Systems for Computer Vision, V.Cantoni, S.Levialdi (eds.), Maratea, 1986 view details
  • Di Gesu, V. "PCL: A Pyramid Language", Tech. Rep. DMAT-2-88, Univ.of Palermo, 1988. view details
          in Proc. NATO ASI Workshop on Pyramidal Systems for Computer Vision, V.Cantoni, S.Levialdi (eds.), Maratea, 1986 view details
  • Chiavetta, F.; Di Gesù, V. ; Gerardi, G.; Tegolo, D. "The Pictorial Language and the Software Environment of the HERMIA Machine" view details
          in 1st Israeli/Italian workshop on Computer Vision, Capri, 1991 view details
  • Di Gesu, Vito; Isgro, Francesco "Contexts and Vision" International Computer Science Institute Technical Report 95-38 view details Extract: PICL description
    Contexts have been firstly introduced in the Pyramid C Language (PCL), as object variables, to handle concurrent processes on the pyramid machine: PAPIA. The PCL is an extension of the C language, where parallel instructions and hierarchical-datastructures, oriented to image analysis computation, are included. PCL allows the execution of asynchronous processes, when the processors of the pyramid layers operate in Multy Single Instruction Multy Data (M-SIMD) mode.
    Contexts variables are also included in the visual language PICL (PIctorial C Language), to program the HERMIA machine (Heterogeneous and Reconfigurable Machine for Image Analysis). The PICL language is an extension of PCL to reconfigurable networks machines. It includes pictorial data and pictorial operators, and reconfiguration network instructions.

    Extract: Conclusions
    Conclusions
    The report describes how concurrent processes are handled in the pictorial language, PICL, designed for image analysis and implemented on the multi-processor machine HERMIA. The concept of context has been introduced, as a hardware-data relation, which is especially useful in the case of distributed spatial data.
    The relation between the open/close statements, formal languages and dependency graphs has been analyzed. The results allow an easy and fast implementation inside PICL. Contexts are handled inside a PICL program as normal dynamic variables, this make easier the handling of concurrent processes and the control of the flow of the program. Moreover, contexts have geometrical and topological properties that allow us to put the computation into the proper data allocation.


          in 1st Israeli/Italian workshop on Computer Vision, Capri, 1991 view details