Image processing apparatus and image processing method

Patent metadata

Field	Value
Publication number	US-2016105679-A1
Application number	US-201414889789-A
Country	US
Kind code	A1
Filing date	Apr 30, 2014
Priority date	May 7, 2013
Publication date	Apr 14, 2016
Grant date	—

Abstract

In imaging processing, image data outputted from an imaging device is acquired as a data string formed of a first display value expressed by a first number of gradations obtained through conversion of a luminance value in accordance with preset nonlinear conversion characteristics. The first display value is compressed in accordance with preset compression characteristics, and outputted as a second display value expressed by a second number of gradations smaller than the first number of gradations. A recognition target is detected from an image expressed by compressed data that is a data string formed of the second display value. A ratio of the second number of gradations to the first number of gradations is taken as a basic compression ratio, a luminance range including at least a recognition target range that is a luminance range where the recognition target is estimated to be present is taken as a specified range, and the first display value corresponding to a boundary luminance value that is a minimum luminance value in the specified range is taken as a boundary first display value. In the recognition target range, the compression characteristics are set so that the second display value is a sum of a compressed value and the boundary first display value, the compressed value being obtained by compressing a value of not less than the boundary first display value among the first display values at a low compression ratio lower than the basic compression ratio.

First claim

1 . An image processing apparatus comprising:
an acquiring means for converting a luminance value in accordance with preset nonlinear conversion characteristics and acquiring, from an imaging device, image data that is a data string formed of a first display value expressed by a first number of gradations; a compressing means for compressing the first display value in accordance with preset compression characteristics and outputting, as a second display value, a value expressed by a second number of gradations smaller than the first number of gradations; and a recognizing means for detecting a recognition target from an image expressed by compressed data that is a data string formed of the second display value, wherein: a ratio of the second number of gradations to the first number of gradations is set as a basic compression ratio, a luminance range including at least a recognition target range that is a luminance range where the recognition target is estimated to be present is set as a specified range, and the first display value corresponding to a boundary luminance value that is a minimum luminance value in the specified range is set as a boundary first display value; the conversion characteristics are set as a combination of linear characteristics that are predetermined for each luminance range; the specified range is set such that the recognition target range includes a changing point luminance value, the changing point luminance value being a luminance value at a changing point of the linear characteristics, the linear characteristics being changed at the changing point in accordance with the conversion characteristics; the compression characteristics are set such that the second display value is a sum of a compressed value and the boundary first display value in the recognition target range, the compressed value being obtained by compressing a value of not less than the boundary first display value among the first display values at a low compression ratio lower than the basic compression ratio; and the compression characteristics are set such that, at least, in the specified range, a relationship of the second display value converted in accordance with total conversion characteristics to the luminance value becomes linear, the total conversion characteristics being set as a combination of the compression characteristics and the conversion characteristics.

2 . The image processing apparatus according to claim 1 , wherein the compression characteristics are set so that the low compression ratio is one in the specified range.

3 - 8 . (canceled)

9 . An image processing method comprising:
an acquiring step of converting a luminance value in accordance with preset nonlinear conversion characteristics and acquiring, from an imaging device, image data that is a data string formed of a first display value expressed by a first number of gradations; a compressing step of compressing the first display value in accordance with preset compression characteristics and outputting, as a second display value, a value expressed by a second number of gradations smaller than the first number of gradations; and a recognizing step of detecting a recognition target from an image expressed by compressed data that is a data string formed of the second display value, wherein: a ratio of the second number of gradations to the first number of gradations is set as a basic compression ratio, a luminance range including at least a recognition target range that is a luminance range where the recognition target is estimated to be present is set as a specified range, and the first display value corresponding to a boundary luminance value that is a minimum luminance value in the specified range is set as a boundary first display value; the conversion characteristics are set as a combination of linear characteristics that are predetermined for each luminance range; the specified range is set such that the recognition target range includes a changing point luminance value, the changing point luminance value being a luminance value at a changing point of the linear characteristics, the linear characteristics being changed at the changing point in accordance with the conversion characteristics; the compression characteristics are set such that the second display value is a sum of a compressed value and the boundary first display value in the recognition target range, the compressed value being obtained by compressing a value of not less than the boundary first display value among the first display values at a low compression ratio lower than the basic compression ratio; and the compression characteristics are set such that, at least, in the specified range, a relationship of the second display value converted in accordance with total conversion characteristics to the luminance value becomes linear, the total conversion characteristics being set as a combination of the compression characteristics and the conversion characteristics.

Assignees

• Denso Corp

Inventors

• Murao Toshikazu
• Kimura Takayuki

Classifications

• H04N19/426 Primary

  using memory downsizing methods · CPC title

• H04N23/61

  based on recognised objects · CPC title

• H04N25/134

  based on three different wavelength filter elements · CPC title

• H04N23/76

  by influencing the image signals · CPC title

• H04N23/71

  Circuitry for evaluating the brightness variation · CPC title