Title: Quality inspection of Sandwich using computer vision
Title: Quality inspection of Sandwich using computer vision
Objective: To build a system to get rating, decision and nutritional values of sandwich.
Scope: This project is limited to quality inspection of BLT sandwich which consists of white bread slices, Bacon, Lettuce and Tomato.
Sequence of ingredients and Preparation of Sandwich:
Bread Slice (Bottom layer)
Bacon
Tomato
Lettuce
Bread Slice (Top layer)
User provided details
Dataset - SVO files and images of ingredients and sandwiches taken by ZED-M Stereo Camera
Nutritional Data of each ingredient in Excel
Material specifications (user inputs)
Ingredient Dimensions (in cm) Quantity Thickness (cm) Area (cm^2) Weight (g) Cost ()
Bread slice 11 x 11 2 1 121 82 0.4
Bacon 11 x 2 5 0.1 110 16 0.2
Lettuce ~ - - 120 20 0.2
Tomato 5 4 0.5 78.5 32 0.2
Total 150 1
Aim of the Project
Detect object (identify each ingredient and label)
Detect quality (abnormalities, colour variations, defects, spoilage etc) each ingredient
find weight using area or size then nutritional values of each ingredient and final sandwich
Check the order of ingredients
Count the pieces of each item
Calculate cost of each ingredient and final sandwich
Rating of Sandwich (= sum of % of fill,
Final Output: Quality rating, cost, nutritional values of sandwich for a given svo or live camera.
Assign score and rate sandwiches based on below quality aspects
Quality Aspect Acceptance Criteria Score (0 or 1) Weightage Result (=Score *Weightage)
Food Quality Uniform colour/texure, no defects or spoilage 1 5 5
Food Quantity Within 10% of targeted Weight or number of slices of each ingredient 0 1 0
Ingredients Present all ingredient in the sandwich 0 1 0
Under/Over filled adequately filled 0 1 0
Size/shape/weight of the sandwich Total weight within 10% of targeted weight or size 0 2 0
Rating (= sum of Result) 5
Decision (Accept if Rating>5, otherwise Reject) Rejected
Note: Score 1 if within acceptance criteria, otherwise 0
Final Output of the code should display for each sandwich
Rating (1 to 10)
Decision (Accepted or Rejected)
Nutritional Values (Energy, Protein, Fat, Saturates, Sugar, Salt)
Example Output for a given sandwich:
Ref:
Detection, quantification and classification of ripened tomatoes: a comparative analysis of image processing and machine learning (wiley.com)Composition of foods integrated dataset (CoFID) - GOV.UK (www.gov.uk)
