Study Of The Application Of Embedded Computer Vision Technology Used In Snowflakes Beef Quality Detection

Snowflakes beef is the hight-quality beef, fat deposition in the muscle have shown a marbling pattern of fat distribution characteristics. The6th to7th ribs eye muscle cross-section of snowflake beef carcass is the most important part of snowflakes beef quality assessment. In the traditional beef production and processing industry, artificial measurement method was used to obtain the information section of the beef eye muscle, which has defect of low efficiency and low consistency effected by personal experience.According snowflake beef actual production China, this paper researched the artificial measurement method and proposed beef quality grading evaluation method based on embedded and machine vision technology to take place the artificial measurement. This project was supported by National Beef Cattle Yak industrial technology system, the actual environment of snowflakes beef and segmentation method was researched and a portable beef quality testing device was developed based on the need of beef production company. In order to improved and solved the problem of previous portable snowflakes beef quality detection terminal device, optimal designed of embedded image acquisition terminal of the LED light source and intelligent camera system, improved the terminal man-machine interface, independent designed power management system provided power protection for the terminal. Digital image processing algorithms for analysis and processing the snowflakes beef image was developed based on snowflakes beef eye muscle section image feature attribute, finally, intelligent snowflake beef quality information detection method was realized with VC++6.0software.The system consists of two main parts:the image collection terminal based on embedded machine vision technology snowflakes beef rib eye section and the server-side based on digital image processing technology intelligent snowflake beef quality information detection.The DevKit8000was used as the Embedded terminal core control platform, a bar LED light was used as embedded machine vision lighting, CCD industrial camera was used for snowflakes beef rib eye section image acquisition, conversion and storage operations, and finally the captured snowflakes beef image was sent to the server to perform image processing and analysis. In addition, based on the Qt/Embedded, this paper developed on human-computer interaction interface in the embedded Linux system environment. The power management system provided power for portable detection terminal equipment, and which meet fully needs of the industry site grading.Snowflakes beef quality information detection server-side used digital image processing technology and image feature extraction method for beef quality intelligent detection. Firstly, the low-pass filtering, median filtering and image pre-processing method were used for image enhancement processing, the method of image segmentation processing have used by OTSU adaptive threshold. And then, the effective eye muscle target area identification was recognised by the regional notation and mathematical morphology method from the binary image. Finally, the snowflakes beef quality information was extracted by the image analysis method, which included the eye muscle color information, effective eye muscle area and contour, marbling area and contour etc. Thus, achieved snowflakes beef quality information intelligent detection.