Army Project Number 2J620901A721 Man-Computer Functions a-00 Distribution of this document is unlimited. i BLANK PAGES IN THIS DOCUMENT WERE NOT FILMED --■" ■ ■» > BESRL Technical Research Reports and Technical Research Notes are intended for sponsors of R&D tasks and other research and military agencies. Any findings ready for implementation at the time of publication are presented in the latter part of the Brief. Upon completion of a major phase of the task, formal recommendations for

more »

... action normally are conveyed to appropriate military agencies by briefing or Disposition Form. ■ L ■----■--FOREWORD The SURVEILLANCE SYSTEMS Project has at its objective the development of scientific research data bearing on the extraction of information from imagery and the products of other sensors, and the efficient storage, retrieval, and transmission of information with an advanced computerized image interpretation facility. Research results an used in future systems design and in the development of enhanced techniques and procedures for all phases of the image interpretation process within the data reduction facility. The MAN-COMPUTER FUNCTIONS Task is one of four research Tasks established in the Support Systems Research Division of BESRL to concentrate on operational segments of the surveillance system. One major effort of the Task is devoted to the development and evaluation of input/output procedures for the description and transmission of tactical intelligence information. The present study concentrated on two work methods of rapid screening to determine the presence or absence of military targets and work methods of setting priorities for detailed interpretation of selected frames of imagery. BRIEF Requirement: To aii«si the effectiveness in the rapid screening of tactical imagery of two techniques for selecting a limited number of frames of high intelligence potential for detailed interpretation and to determine the effects of time allowed per frame on the effectiveness of the two techniques. Procedure: Two samples of image interpreters (N's of 33 and 30), each sample consisting of three matched groups, screened three sets of imagery at three display time intervals-S, 15, and 25 seonds per frame for Sample 1, and 10, 20, and 30 seconds per frame for Sample 2. Subjects scanned each print for the prescribed time while performing two screening functions: (1) annotating on the frame all areas of military activity detected, and (2) assigning priority ratings of High, Medium, and Law to indicate the estimated intelligence value of the frame. The two methods were compared in terms of accuracy rate and validity at the different display intervals.. Findings: Priority ratings were better indexes of the information potential of the frames than were the annotations. The ratings, of high accuracy even with short viewing time, improved both in accuracy and in validity with longer display time. The validity of the ratings was higher with sets of imagery characterized by relatively few target areas and less complex background. The validity of the number of annotations on a frame, generally low, did not increase with increased display time. More incorrect as well as correct annotations were made. Utilization of Findings: The priority rating method of screening is sufficiently accurate to be useful even where only very short screening time can be allowed. However, imagery differing in complexity may require adjustments in screening time. t^^MB^B After 25 seconds have elapsed, say: READY? BEG Hi! STOPI TAKE YOUR NEXT IRINT. After 25 seconds have elapsed, say: STOP! Continue this procedure until all fifteen prints have been ccqpleted; then have them replace the prints In .he packet. -19