System for food recognition method using portable devices having digital cameras
First Claim
1. A method for food recognition using a portable device equipped with a digital camera to record food and calories, the method comprising:
- segmenting items of food of a top-view photo of a plate of food into regions using the centers, selected by a user, of each item of food;
broadening the regions based on each center and on different color spaces (HSV, RGB, and LAB);
providing three hypotheses of segments chosen for each center to obtain robust results from the segmentation;
training, for each hypothesis, a Support Vector Machine (SVM) having a radial based kernel-type function;
obtaining, from a predetermined menu of food images, a list of probabilities for each segment of food;
choosing the segment with the highest probability; and
showing the results of segmented food and their possible identifications.
2 Assignments
0 Petitions
Accused Products
Abstract
A method for automatic food recognition by using portable devices equipped with digital cameras. With this system, it is possible to identify a previously established food menu. To this purpose, a semi-automated method of segmentation is applied to delineate the regions in which each type of food in an image of a plate of food, captured by a user. Pattern recognition techniques are used in images, integrated into a system whose goal is to label each type of food contained in the photo of a plate of food. No type of preprocessing is performed to correct deficiencies in capturing the image, just using the auto-focus component present in the portable device to capture a clear image.
-
Citations
20 Claims
-
1. A method for food recognition using a portable device equipped with a digital camera to record food and calories, the method comprising:
-
segmenting items of food of a top-view photo of a plate of food into regions using the centers, selected by a user, of each item of food; broadening the regions based on each center and on different color spaces (HSV, RGB, and LAB); providing three hypotheses of segments chosen for each center to obtain robust results from the segmentation; training, for each hypothesis, a Support Vector Machine (SVM) having a radial based kernel-type function; obtaining, from a predetermined menu of food images, a list of probabilities for each segment of food; choosing the segment with the highest probability; and showing the results of segmented food and their possible identifications. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20)
-
Specification