Object detection apparatus

• US 9,418,287 B2
• Filed: 03/11/2014
• Issued: 08/16/2016
• Est. Priority Date: 03/13/2013
• Status: Active Grant

• Current Assignee: DENSO Corporation
• Original Assignee: DENSO Corporation
• Inventors: Kamiya, Yasunori
• Primary Examiner(s): Chu, Randolph I
• Application Number:
  US14/204,152
• Publication Number:
  US 20140270548A1
• Time in Patent Office: 889 Days
• Field of Search: None
• US Class Current: 1/1
• CPC Class Codes: G06F 18/217 Validation; Performance evaluation; Active pattern learning techniques
• CPC Class Codes: G06F 18/285 Selection of pattern recognition techniques, e.g. of classifiers in a multi-classifier system
• CPC Class Codes: G06F 2218/12 Classification; Matching
• CPC Class Codes: G06V 10/776 Validation; Performance evaluation
• CPC Class Codes: G06V 10/87 using selection of the recognition techniques, e.g. of a classifier in a multiple classifier system
• CPC Class Codes: G06V 20/56 exterior to a vehicle by using sensors mounted on the vehicle

Abstract

An object detection apparatus mounted in a system for detecting a target object in various changing environmental conditions. In the apparatus, an acquirer acquires either or both of information indicative of an external environment around the system and information indicative of an operating state of the system. A determiner determines an environmental condition with respect to an input image according to the acquired information. A setter sets, for each of plural image recognition methods each being a combination of one of the plural image recognition dictionaries and one of the plural image recognition techniques, a weighting factor according to the environmental condition determined by the determiner. A detector detects the object in the input image by applying each of the plural image recognition methods to the input image to obtain recognition results reflecting the weighting factors, and collectively evaluating the recognition results.

Claims

1. An object detection apparatus mounted in a system, the apparatus comprising:

a storage storing plural image recognition dictionaries each describing reference data for an object to be detected and plural image recognition techniques each used to detect the object from an input image captured by an imaging unit with use of one of the plural image recognition dictionaries;

an acquirer configured to acquire either or both of information indicative of an external environment around the system and information indicative of an operating state of the system;

a determiner configured to determine an environmental condition with respect to the input image according to the information acquired from the acquirer;

a setter configured to set, for each of plural image recognition methods predefined for respectively different assumed environmental conditions, a weighting factor according to the environmental condition determined by the determiner, each image recognition method being a combination of one of the plural image recognition dictionaries and one of the plural image recognition techniques; and

a detector configured to detect the object in the input image by applying each of the plural image recognition methods to the input image to obtain image recognition results reflecting the weighting factors set by the setter for the respective image recognition methods, and then collectively evaluating the image recognition results.

2. The apparatus of claim 1, wherein the setter is configured to set, for each of the plural image recognition methods, a weighting factor as a function of a similarity between the environmental condition determined by the determiner and the assumed environmental condition corresponding the image recognition method.

3. The apparatus of claim 1, wherein each of the plural image recognition techniques comprises an image recognition algorithm to be applied to the input image with use of the image recognition dictionary of the image recognition method corresponding to the image recognition technique.

4. The apparatus of claim 3, wherein at least one of the plural image recognition techniques further comprises image correction applied to the input image prior to application of the corresponding image recognition algorithm to the input image.

5. The apparatus of claim 1, wherein

the system is a vehicle,

the information indicative of an external environment around the system includes at least one of weather conditions, times of day, ambient brightness, and backdrop attributes, and

the information indicative of an operating state of the system includes at least one of operating states of lighting devices of the vehicle and operating states of a wiper of the vehicle.

6. An object detection apparatus mounted in a system, the apparatus comprising:

a storage storing plural image recognition dictionaries each describing reference data for an object to be detected, and a single image recognition technique used to detect the object from an input image captured by an imaging unit with use of one of the plural image recognition dictionaries;

an acquirer configured to acquire either or both of information indicative of an external environment around the system and information indicative of an operating state of the system;

a determiner configured to determine an environmental condition with respect to the input image according to the information acquired from the acquirer;

a setter configured to set, for each of plural image recognition methods predefined for respectively different assumed environmental conditions, a weighting factor according to the environmental condition determined by the determiner, each image recognition method being a combination of one of the plural image recognition dictionaries and the single image recognition technique; and

a detector configured to detect the object in the input image by applying each of the plural image recognition methods to the input image to obtain image recognition results reflecting the weighting factors set by the setter for the respective image recognition methods, and then collectively evaluating the image recognition results.

7. The apparatus of claim 6, wherein the setter is configured to set, for each of the plural image recognition methods, a weighting factor as a function of a similarity between the environmental condition determined by the determiner and the assumed environmental condition corresponding the image recognition method.

8. The apparatus of claim 6, wherein the single image recognition technique comprises an image recognition algorithm to be applied to the input image with use of one of the image recognition dictionaries.

9. The apparatus of claim 8, wherein the single image recognition technique further comprises image correction applied to the input image prior to application of the image recognition algorithm to the input image.

10. The apparatus of claim 6, wherein

the system is a vehicle,

the information indicative of an external environment around the system includes at least one of weather conditions, times of day, ambient brightness, and backdrop attributes, and

the information indicative of an operating state of the system includes at least one of operating states of lighting devices of the vehicle and operating states of a wiper of the vehicle.