Intelligent Computer Vision System (saif) for Automated Inspection of Ginseng Roots Quality

Intelligent computer-vision system for automated inspection of food safety and quality (SAIF) developed on the basis of compact CCD camera with IEEE-1396 interface and configurable software (IMAQ TM 6.1, Lab VIEW 7.0) is presented. It offers an extensive set of optimized functions for advanced image acquisition, segmentation, feature extraction, data analysis, spatial measurement and calibration. It also includes the ability to set up complex pass/fail decisions in order to control digital I/O devices such as PLC. The system application for online inspection of ginseng root quality during drying was developed. Area shrinkage was continuously monitored through computer-vision system by extracting morphological features with thresholding and pixels counting. Colour changes were monitored through computer-vision system as surface color intensity. Relationships between image attributes and physical parameters of drying (shrinkage/moisture, color/quality) were used for online estimation of actual moisture content and quality degradation. Testing of system proved accuracy in estimation of ginseng quality and process parameters in multi-stage drying. The feasibility of SAIF as system observer for closed-loop control is discussed.