An official website of the United States government.

Official websites use .gov
A .gov website belongs to an official government organization in the United States.

Secure .gov websites use HTTPS
A lock ( ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites.



Goal: The goal of this project is to adapt SC's existing image processing technology for visual inspection of meat during processing and efficient classification by fat content. Visual inspection of meat to estimate fat content is one of the steps in processing at small and medium sized companies where the associated inaccuracy results in process inefficiency, wastage, and potential errors, leading to higher production costs and reduced profitability. There is a significant need and substantial growth potential in the small to medium sized meat processing industry for an inexpensive tool for real-time detection and classification of lean-to-fat ratios and other meat attributes such as the size of the cut. SC will work closely with our meat processing partner, Lorentz Meats, Cannon Falls, MN to customize the technology for their meat processing environment. The technology will be applicable to smart processing and quality control (QC) of meat and will be designed to be HAACP and HIMP compliant.SC and its collaborators are committed to test and validate the technical and commercial viability of our solution as early adopters with this pilot program. SC expects that the successful completion of Phase III will secure at least one client. Larta and SC have identified more prospective small to medium-sized meat and poultry processors (MPP's)in the Midwest who have shown strong interest in adopting this technology. After the pilot test at the initial site, the pilot data and analysis results will be packaged into sales and marketing materials, used for pitching to other prospective clients.Objectives: The objectives are:The meat inspection tool (MIT) prototype employing optical image analysis and machine learning (ML) algorithms will be able to estimate fat content of the meat to within 3% accuracy.MIT will use commercial off-the-shelf (COTS) components (system hardware will be designed to cost less than $35K).Fat content estimation will be performed in real-time and in-line with no interruption or disruption to the work process, or use of floor space.The tool will have a small footprint so it may be seamlessly deployed in the processing plant.The effectiveness of the tool will be demonstrated at the processing plant of our partner, Lorentz Meats.Development of pilot-specific and broad (non-pilot) strategies, analysis, planning and execution. These include demonstrating improvement in product quality, possibly increase productivity and therefore profits and assure meeting USDA regulatory requirements.SC will work with Larta to develop test protocol, milestone objectives, measurable goals, and a business model that may be a baseline model for all future SC's customer pursuits in the small to medium sized meat and poultry processing industry.

Ghosal, S.; Ebert, JO, .; de Roover, DI, .; Porter, LA, .; Emami-Naeini, AB, .
Start date
End date
Project number
Accession number