IMAGE RECOGNITION DEVICE, IMAGE RECOGNITION METHOD, AND IMAGE RECOGNITION PROGRAM

• US 20130279745A1
• Filed: 01/31/2013
• Published: 10/24/2013
• Est. Priority Date: 02/01/2012
• Status: Active Grant

• Current Assignee: Honda Elesys Co., Ltd. (Nidec Corporation)
• Original Assignee: C/O Honda Elesys Company Limited
• Inventors: KANAMOTO, Junji, MITSUI, Tomokazu, AKIMOTO, Hiroyuki
• Granted Patent: US 9,064,182 B2
• Time in Patent Office: Days
• US Class Current: 382/103
• CPC Class Codes: G06V 10/446 using Haar-like filters, e.g. using integral image techniques
• CPC Class Codes: G06V 10/462 Salient features, e.g. scale invariant feature transforms [SIFT]

Abstract

An image recognition device includes an image acquiring unit configured to acquire an image, and an object recognition unit configured to calculate gradient directions and gradient values of intensity of the image acquired by the image acquiring unit, to scan the gradient values of each acquired gradient direction with a window, to calculate a rectangular feature value, and to recognize a target object using a classifier based on the rectangular feature value.

Claims

1. An image recognition device comprising:

an image acquiring unit configured to acquire an image; and

an object recognition unit configured to calculate gradient directions and gradient values of intensity of the image acquired by the image acquiring unit, to scan the gradient values of each acquired gradient direction with a window, to calculate a rectangular feature value, and to recognize a target object using a classifier based on the rectangular feature value.

2. The image recognition device according to claim 1, wherein the object recognition unit uses as the rectangular feature value one or more kinds of:

a single-rectangle feature value;

a Haar-like feature value;

feature values based on a plurality of features (Haar-like application) with different rectangular areas adjacent to each other in the same gradient direction;

feature values based on a plurality of features with equal or different rectangular areas separated in the same gradient direction; and

feature values based on a plurality of features with equal or different rectangular areas separated in the different gradient directions.

3. The image recognition device according to claim 1, wherein the object recognition unit applies any operation of four arithmetic operations to the rectangular feature values when relationships of a plurality of rectangles are used as feature values.

4. The image recognition device according to claim 1, wherein the object recognition unit performs normalization based on an illumination difference using as the rectangular feature value any one of:

an average value of a window unit;

a standard deviation value of the window unit;

an average value of a raster scan region; and

a standard deviation value of the raster scan region.

5. The image recognition device according to claim 1, wherein the object recognition unit selects a weak classifier obtained through learning of rectangular features as a classifier and uses the selected weak classifier for recognition.

6. The image recognition device according to claim 1, wherein the object recognition unit creates the classifier through boosting or other ensemble learning.

7. The image recognition device according to claim 1, wherein the object recognition unit uses an AdaBoost classifier or a real AdaBoost classifier as the classifier.

8. The image recognition device according to claim 1, wherein the object recognition unit uses coefficients for normalizing a range of the rectangular feature values to select the classifier.

9. The image recognition device according to claim 8, wherein the coefficients for normalizing the range of rectangular feature values are values simultaneously learned when creating the classifier through learning.

10. An image recognition method comprising:

causing an image acquiring unit to acquire an image; and

causing an object recognition unit to calculate gradient directions and gradient values of intensity of the image acquired by the image acquiring unit, to scan the gradient values of each acquired gradient direction with a window, to calculate a rectangular feature value, and to recognize a target object using a classifier based on the rectangular feature value.

11. An image recognition program causing a computer to perform:

a sequence of causing an image acquiring unit to acquire an image; and

a sequence of causing an object recognition unit to calculate gradient directions and gradient values of intensity of the image acquired by the image acquiring unit, to scan the gradient values of each acquired gradient direction with a window, to calculate a rectangular feature value, and to recognize a target object using a classifier based on the rectangular feature value.