CN106791476A - A kind of image-pickup method and device - Google Patents

Abstract

The invention discloses an image acquisition method and device to solve the technical problem in the prior art that the area of interest in the image is incorrectly exposed when the image is acquired. The method includes: determining a region of interest ROI from an image collected by a camera according to a preset target object type; calculating an exposure evaluation index of the ROI, and calculating an exposure reference according to the exposure evaluation index; and the preset exposure standard value, adjust the camera parameters of the camera; collect the next frame of image according to the camera parameters. Adopt the technical scheme of the present invention. The technical problems existing in the prior art are overcome, and the exposure accuracy of the region of interest in the image is improved.

Description

An image acquisition method and device

technical field

The invention relates to the field of computers, in particular to an image acquisition method and device.

Background technique

Currently, the camera's automatic adjustment algorithm is based on various artificial assumptions, such as global average metering, center-weighted metering, spot metering, etc. However, in scenarios such as advanced assisted driving or automatic driving, the environment changes rapidly and the environment differs greatly during the process of image acquisition by the camera. It is difficult to automatically adjust the camera using the same criterion during the entire image acquisition process; in addition, in some specific In the application, the importance of different regions in the same image is also inconsistent. For example, in automatic driving, more attention should be paid to the road surface, vehicles on the road, pedestrians, tricycles and other regions of interest (ie ROI, Region of Interest), and not Do not pay attention to areas such as the sky; for example, in security monitoring, pay more attention to the face area. However, the existing camera automatic adjustment algorithm is not well adapted to these specific applications, resulting in incorrect exposure of the ROI, making the collected image effect poor, and thus making subsequent processing based on the collected image less accurate. Performance is poor.

Contents of the invention

In view of the above problems, the present invention provides an image acquisition method and device to solve the technical problem in the prior art that the poor automatic adjustment algorithm leads to incorrect exposure of the image acquired by the camera.

Embodiments of the present invention, on the one hand, provide an image acquisition method, the method includes:

Determine the region of interest ROI from the image collected by the camera according to the preset target object type;

calculating an exposure evaluation index of the ROI, and calculating an exposure benchmark according to the exposure evaluation index;

Adjusting camera parameters of the camera according to the exposure reference and the preset exposure standard value;

Acquire the next frame of image according to the camera parameters.

Embodiments of the present invention, on the other hand, provide an image acquisition device, the device includes:

A region determining unit, configured to determine the region of interest ROI from the image collected by the camera according to the preset target object type;

a calculation unit, configured to calculate an exposure evaluation index of the ROI, and calculate an exposure benchmark according to the exposure evaluation index;

an adjustment unit, configured to adjust camera parameters of the camera according to the exposure reference and a preset exposure standard value;

The acquisition unit is configured to acquire the next frame of image according to the camera parameters.

In the technical solution of the present invention, the ROI in the image is determined according to the preset target object type, and the exposure reference is calculated according to the exposure evaluation index of the ROI, and the camera parameters of the camera are adjusted based on the exposure reference and the preset exposure standard value , using the adjusted camera parameters to capture the next frame of image. Due to the high frequency of image acquisition by the camera and the short time interval, there is little difference in the scene captured between the previous frame image and the next frame image, and the ROI overlap between the previous frame image and the next frame image is large. Therefore, The exposure reference calculated based on the exposure evaluation index of the ROI of the previous frame image is more accurate and more in line with the ROI exposure requirements of the next frame image. Therefore, the camera parameters are adjusted based on the exposure reference and the preset exposure standard value. More accurately, the next frame of image is collected with the adjusted camera parameters, so that the ROI exposure effect in the next frame of image is better.

Additional features and advantages of the invention will be set forth in the description which follows, and in part will be apparent from the description, or may be learned by practice of the invention. The objectives and other advantages of the invention may be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings.

The technical solution of the present invention will be described in further detail below with reference to the drawings and embodiments.

Description of drawings

The accompanying drawings are used to provide a further understanding of the present invention, and constitute a part of the description, and are used together with the embodiments of the present invention to explain the present invention, and do not constitute a limitation to the present invention.

Fig. 1 is the method flowchart of image acquisition method in the embodiment of the present invention;

Fig. 2 is a schematic structural diagram of an image acquisition device in an embodiment of the present invention.

detailed description

In order to enable those skilled in the art to better understand the technical solutions in the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described The embodiments are only some of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts shall fall within the protection scope of the present invention.

Embodiment one

Referring to Fig. 1, it is a method flowchart of an image acquisition method in an embodiment of the present invention, the method includes:

Step 101 , according to the preset target object type, determine the ROI from the image collected by the camera.

In the embodiment of the present invention, the target object type can be flexibly set according to the application scenario, which is not strictly limited in this application. For example, in the field of automatic driving, the target object type can be one or a combination of the following: road surface, traffic signs, pedestrians, vehicles, tricycles, motorcycles, electric vehicles, etc.; in the security field, the target object type can be human face, Pedestrians and so on.

Preferably, in step 101, the ROI is determined from the image collected by the camera. For specific implementation, please refer to step A-step B:

Step A, using a preset algorithm (such as target detection algorithm, semantic segmentation algorithm, scene segmentation algorithm, etc., which is not strictly limited in this application) to determine the semantic label of each pixel in the image;

In the embodiment of the present invention, the types of semantic tags correspond to the types of target objects. For example, in the field of automatic driving, the types of semantic tags include one or more of the following: road surfaces, traffic signs, vehicles, bicycles, tricycles, motorcycles, Electric vehicles, etc. In the field of security, semantic tags include human faces or pedestrians, etc.

Step B. Determine the ROI in the image according to the pixels whose semantic labels match the type of the target object.

Preferably, the aforementioned step B can be achieved in the following two ways, but not limited to:

Mode 1. Determine the closed area formed by the pixels whose semantic label matches the type of the target object as the ROI in the image. Among them, the closed region contains pixels whose semantic labels match the target object type.

Mode 2: Determine a set of pixels whose semantic labels match the type of the target object as an ROI in the image.

Step 102, calculating an exposure evaluation index of the ROI, and calculating an exposure reference according to the exposure evaluation index.

The specific implementation of step 102 may include but not limited to the following methods: calculating the exposure histogram of the pixels included in the ROI; calculating the median or average of the exposure histogram, and determining the calculated median or average as Exposure benchmark.

Those skilled in the art can also obtain the exposure benchmark by means of counting the partition histogram of the ROI area, the category histogram, the mean variance of the regional brightness, etc., which will not be described in detail here.

Step 103 , adjusting camera parameters of the camera according to the exposure reference and a preset exposure standard value.

In the embodiment of the present invention, the exposure standard value is determined in advance according to the application scenario, and is an empirical value, which is not strictly limited in this solution. For example, in an automatic driving scenario, if the target object type is a vehicle and/or a road surface, the corresponding exposure standard value is set to 90.

Preferably, in step 103 of the embodiment of the present invention, it can be realized in the following manner, but not limited to: input the exposure reference and exposure standard value into a preset PID (proportional-integral-derivative controller), and the PID controls The camera parameters of the camera.

In the embodiment of the present invention, the camera parameters may include shutter time, aperture, ISO (that is, light sensitivity) value, and the like.

Step 104. Acquire the next frame of image according to the camera parameters.

Embodiment two

Referring to Fig. 2, it is a schematic diagram of the mechanism of the image acquisition device in the embodiment of the present invention, the device includes:

A region determining unit 21, configured to determine the region of interest ROI from the image collected by the camera according to the preset target object type;

A calculation unit 22, configured to calculate an exposure evaluation index of the ROI, and calculate an exposure reference according to the exposure evaluation index;

An adjustment unit 23, configured to adjust camera parameters of the camera according to the exposure reference and a preset exposure standard value;

The acquisition unit 24 is configured to acquire the next frame of image according to the camera parameters.

Preferably, the region determination unit 21 is specifically configured to: use a preset algorithm to determine the semantic label of each pixel in the image; determine the ROI in the image according to the pixels whose semantic label matches the type of the target object .

Preferably, the region determination unit 21 determines the ROI in the image according to the pixels whose semantic labels match the type of the target object, specifically for:

Determining a closed area formed by pixels whose semantic labels match the type of the target object as the ROI in the image; or, determining a set of pixels whose semantic labels match the type of the target object as the ROI in the image ROIs in the image.

Preferably, the computing unit 22 specifically includes:

a first calculation subunit, configured to calculate an exposure histogram of pixels included in the ROI;

The second calculation subunit is used to calculate the median or average of the exposure histogram, and determine the calculated median or average as the exposure reference.

Preferably, the adjustment unit 23 is specifically configured to: input the exposure reference and the exposure standard value into a proportional-integral-derivative controller PID, and the PID controls the camera parameters of the camera.

In the technical solution of the present invention, the ROI in the image is determined according to the preset target object type, and the exposure reference is calculated according to the exposure evaluation index of the ROI, and the camera parameters of the camera are adjusted based on the exposure reference and the preset exposure standard value , using the adjusted camera parameters to capture the next frame of image. Due to the high frequency of image acquisition by the camera and the short time interval, there is little difference in the scene captured between the previous frame image and the next frame image, and the ROI overlap between the previous frame image and the next frame image is large. Therefore, The exposure reference calculated based on the exposure evaluation index of the ROI of the previous frame image is more accurate and more in line with the ROI exposure requirements of the next frame image. Therefore, the camera parameters are adjusted based on the exposure reference and the preset exposure standard value. More accurately, the next frame of image is collected with the adjusted camera parameters, so that the ROI exposure effect in the next frame of image is better.

The above is the core idea of the present invention. In order to enable those skilled in the art to better understand the technical solutions in the embodiments of the present invention, and to make the above-mentioned purposes, features and advantages of the embodiments of the present invention more obvious and understandable, the following is combined with the attached The figure further explains in detail the technical solutions in the embodiments of the present invention.

Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention also intends to include these modifications and variations.

Claims (10)

1. a kind of image-pickup method, it is characterised in that including：

According to preset destination object type, region of interest ROI is determined from the image of camera collection；

The exposure evaluation index of the ROI is calculated, and Expose f iotaducials are calculated according to the exposure evaluation index；

According to the Expose f iotaducials and preset exposure standard value, the camera parameter of the camera is adjusted；

Next two field picture is gathered according to the camera parameter.

2. method according to claim 1, it is characterised in that according to preset destination object type, from camera collection ROI is determined in image, is specifically included：

The semantic label of each pixel in described image is determined using preset algorithm；

According to semantic label and the pixel of the destination object type matching, the ROI in described image is determined.

3. method as claimed in claim 2, it is characterised in that according to the picture of semantic label and the destination object type matching Element, determines the ROI in described image, specifically includes：

The closed area that institute's semantic tags are constituted with the pixel of the destination object type matching is defined as in described image ROI；

Or, the set that institute's semantic tags are constituted with the pixel of the destination object type matching is defined as in described image ROI.

4. the method for claim 1, it is characterised in that calculate the exposure evaluation index of the ROI, and according to the exposure Light evaluation index calculates Expose f iotaducials, specifically includes：

Calculate the exposure histogram of the pixel that the ROI is included；

The histogrammic median of exposure or average are calculated, the median or average that will be calculated are defined as exposing base It is accurate.

5. the method for claim 1, it is characterised in that according to the Expose f iotaducials and preset exposure standard value, adjusts The camera parameter of the whole camera, specifically includes：

The Expose f iotaducials and exposure standard value are input into proportional-integral derivative controller PID, as described in the PID control The camera parameter of camera.

6. a kind of image collecting device, it is characterised in that including：

Area determination unit, for according to preset destination object type, region of interest being determined from the image of camera collection Domain ROI；

Computing unit, the exposure evaluation index for calculating the ROI, and exposure base is calculated according to the exposure evaluation index It is accurate；

Adjustment unit, for according to the Expose f iotaducials and preset exposure standard value, adjusting the camera parameter of the camera；

Collecting unit, for gathering next two field picture according to the camera parameter.

7. device as claimed in claim 6, it is characterised in that area determination unit, specifically for：

The semantic label of each pixel in described image is determined using preset algorithm；

According to semantic label and the pixel of the destination object type matching, the ROI in described image is determined.

8. device as claimed in claim 7, it is characterised in that area determination unit is according to semantic label and the destination object The pixel of type matching, determines the ROI in described image, specifically for：

The closed area that institute's semantic tags are constituted with the pixel of the destination object type matching is defined as in described image ROI；

Or, the set that institute's semantic tags are constituted with the pixel of the destination object type matching is defined as in described image ROI.

9. the device as described in right wants 6, it is characterised in that computing unit, specifically includes：

First computation subunit, the exposure histogram for calculating the pixel that the ROI is included；

Second computation subunit, for calculating the histogrammic median of exposure or average, the median that will be calculated Or average is defined as Expose f iotaducials.

10. device as claimed in claim 6, it is characterised in that the adjustment unit, specifically for：By the Expose f iotaducials It is input into proportional-integral derivative controller PID with exposure standard value, the camera parameter of camera as described in the PID control.