Unlock instant, AI-driven research and patent intelligence for your innovation.

Image processing device and image processing method

An image processing device and technology of brightness value, which are applied in image data processing, image enhancement, image analysis, etc.

Active Publication Date: 2017-10-10
DENSO CORP
View PDF1 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

In other words, in the case where the coarse resolution region includes the luminance of the recognition object, the detection accuracy of the recognition object may decrease

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Image processing device and image processing method
  • Image processing device and image processing method
  • Image processing device and image processing method

Examples

Experimental program
Comparison scheme
Effect test

no. 1 Embodiment approach

[0027] The image processing device of the present embodiment is applied to a driving assistance system mounted on a vehicle, which detects a recognized object existing in front of the vehicle (own vehicle) and executes various driving assistance controls based on the detection result.

[0028] driver assistance system 1 such as figure 1 As shown, a vehicle-mounted camera (hereinafter, simply referred to as a camera) 20 is provided, which captures the front of the vehicle VE and outputs the captured image as 12-bit image data. This camera functions as a photographing unit. In addition, the driving assistance system 1 includes an image processing device 10 that converts the image data output from the camera 20 into 8-bit image data, and executes various functions related to the image represented by the converted image data. Image processing corresponding to the purpose (application) (for example, the processing of detecting recognition objects such as vehicles ahead, pedestrian...

Deformed example 1

[0078] In the above-described embodiment, the low compression ratio β is set to 1 according to the compression characteristics, but the setting of the low compression ratio β is not limited to this. The low compression ratio β may be set to a compression ratio lower than the basic compression ratio. In this case (β≠1), the compressed pixel value Z is 2 8 When the luminance value Le becomes Le>L4, the dynamic range can be expanded.

[0079] In addition, in the above-described embodiment, the same low compression ratio β as the specified range is set in the luminance range larger than the specified range according to the compression characteristics, but the present invention is not limited thereto. For a luminance range larger than the specified range, a compression ratio higher than the low compression ratio β that is different from the low compression ratio β may be set. As a result, the contrast between the white line and the road surface can be made clearer, and the dynami...

Deformed example 2

[0082] The compression characteristic can also be set to include the change point luminance value L of the HDR characteristic in the recognition target range 1 case (for example Figure 7 As shown, when the object of recognition is a gray vehicle for the HDR characteristic), the relationship between the compressed pixel value and the luminance value becomes linear within a specified range through transformation based on the comprehensive transformation characteristic.

[0083] Specifically as Figure 8 As shown, the brightness range of the gray vehicle (recognition object range) can also be the brightness value L 5 ~ L 6 case, the brightness range L 5 ~ L 6 As a specified range, the compression characteristic is set such that the relationship of the compressed pixel value with respect to the luminance value becomes linear within the specified range.

[0084] Thus, at least for the specified range L 5 ~ L 6 , since the compressed pixel value has a grayscale difference th...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

According to the image processing device and the image processing method, the imaging data output by the imaging element is obtained as a data string composed of the first display value represented by the first gray scale number by converting the luminance value with a preset nonlinear transformation characteristic. (step S110). The first display value is compressed with a preset compression characteristic, and a value represented by a second gradation number smaller than the first gradation number is output as a second display value (step S120 ). A recognition object is detected from an image represented by compressed data which is a data string composed of the second display value (step S130). Here, the ratio of the second gradation number to the first gradation number is taken as the basic compression ratio, and the luminance range including at least the luminance range estimated to exist in the recognition target, that is, the recognition target range, is taken as the specified range, and the specified The above-mentioned first display value corresponding to the minimum brightness value in the range, that is, the boundary brightness value is used as the boundary first display value. The compression characteristic is set so that the second display value becomes a value obtained by compressing values ​​above the boundary first display value among the first display values ​​at a low compression ratio lower than the basic compression ratio in the recognition target range, which is the same as the boundary first display value. sum of values.

Description

technical field [0001] The present invention relates to an image processing device and an image processing method for detecting a recognition object from an image captured by a camera such as a vehicle-mounted camera. Background technique [0002] Conventionally, there is known an image processing device that detects a preset recognition target from an image captured by a camera. As this image processing device, devices with various processing capabilities are used depending on the application. The camera outputs image data with a wide dynamic range due to improved performance. Therefore, there is an image processing device that, when the dynamic range (for example, 12 bits) of the image data output from the camera exceeds the processing capability (for example, 8 bits) of the image processing device, the image output from the camera The data undergoes grayscale transformation (for example, conversion from 12 bits to 8 bits) to reduce the number of grayscales, and the imag...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Patents(China)
IPC IPC(8): G06T5/00H04N1/407H04N5/225B60R1/00
CPCG06T2207/30252H04N23/61H04N23/71H04N23/76H04N25/134G06T5/92H04N19/426
Inventor 村尾俊和木村贵之
Owner DENSO CORP