The desire to transmit high-quality, single-frame images such as reconnaissance photographs places a severe strain on narrow-band communications channels. The problem is severe enough to warrant use of complex and expensive compression methods, providing they really work. The authors propose a method of evaluating compression algorithms applied to pictures, and apply it to several well-known techniques already having one or more hardware implementations. The importance of proper data management for pictorial communication is illustrated, and by implication are shown to be much more important than the data compression algorithms. Relevant to the data management proposals, some methods of improving the cosmetic quality of sparsely sampled images are shown. An interactive computer system (AFCRL's Image Analysis Facility) was used in the study. Special features of this facility are operator control of the entire process from image acquisition through final display, interactive processing, and the ready ability to explore new ideas in some depth. (Author).

The Image Analysis Facility (IAF) is a special-purpose digital computer system incorporating custom hardware and software that has been developing at the AFCRL Data Sciences Laboratory for over 10 years. The IAF is designed to provide highly flexible, on-line interactive processing of pictorial information that includes capabilities that are unique to a non-batch computer system. The IAF system hardware and software sections are broken down functionally and are fully discussed. Over 40 photographs illustrating selected program effects on pictures are presented. (Author).

The desire to transmit high-quality, single-frame images such as reconnaissance photographs places a severe strain on narrow-band communications channels. The problem is severe enough to warrant use of complex and expensive compression methods, providing they really work. The authors propose a method of evaluating compression algorithms applied to pictures, and apply it to several well-known techniques already having one or more hardware implementations. The importance of proper data management for pictorial communication is illustrated, and by implication are shown to be much more important than the data compression algorithms. Relevant to the data management proposals, some methods of improving the cosmetic quality of sparsely sampled images are shown. An interactive computer system (AFCRL's Image Analysis Facility) was used in the study. Special features of this facility are operator control of the entire process from image acquisition through final display, interactive processing, and the ready ability to explore new ideas in some depth. (Author)