CN107564053B - Parallax value correction method and device applied to road scene - Google Patents
Parallax value correction method and device applied to road scene Download PDFInfo
- Publication number
- CN107564053B CN107564053B CN201710761404.6A CN201710761404A CN107564053B CN 107564053 B CN107564053 B CN 107564053B CN 201710761404 A CN201710761404 A CN 201710761404A CN 107564053 B CN107564053 B CN 107564053B
- Authority
- CN
- China
- Prior art keywords
- value
- pixel
- parallax
- sub
- segment
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Abstract
The invention provides a parallax value correction method and a device applied to a road scene, wherein the method comprises the following steps: calculating a disparity map of the reference map according to the reference map and the matching map of the same road scene; dividing pixel points of each line in a road area of the reference image into at least 2 pixel sections; dividing the pixel segment into at least 2 sub-pixel segments; if the absolute value of the difference value of the parallax mean value of the sub-pixel segment and the pixel segment to which the sub-pixel segment belongs is larger than a preset third parallax threshold, determining that the sub-pixel segment is a sub-pixel segment with an abnormal parallax value; and correcting the parallax value of the pixel point of the sub-pixel segment with the abnormal parallax value according to the parallax value of other sub-pixel segments of which the absolute value of the difference value between the adjacent parallax average value and the parallax average value of the pixel segment to which the adjacent parallax average value belongs is smaller than a preset third parallax threshold value, so as to improve the accuracy of the parallax value of the road area.
Description
Technical Field
The invention relates to the technical field of computer vision, in particular to a parallax value correction method and device applied to a road scene.
Background
The binocular stereo vision technology is a technology for performing parallax estimation on a depth map of the same scene shot by a binocular camera to determine the depth of the scene in a three-dimensional space. Since the parallax estimation technology is not mature, the accuracy of the estimated parallax value is not high, and in many cases, a corresponding parallax value correction scheme needs to be formulated for a specific application scene to improve the accuracy of the parallax value.
In the auxiliary driving application, the road condition detection based on the road disparity map is a very important function, and when poor road conditions are detected, the safety driving can be favorably ensured through early warning. However, since a part of the road surface area is relatively flat and texture information is less, the accuracy of parallax estimation is not high, a large number of false alarms are generated, and driving smoothness is affected. At present, no scheme for correcting the parallax value in a road scene exists.
Disclosure of Invention
The invention aims to provide a parallax value correction method and a parallax value correction device applied to a road scene, which are used for improving the accuracy of the parallax value in the road scene.
In order to achieve the purpose, the invention provides the following technical scheme:
in one aspect, the present invention provides a method for correcting a disparity value applied in a road scene, where the method includes:
calculating a disparity map of the reference map according to the reference map and the matching map of the same road scene;
dividing pixel points in each row in a road area of the reference image into at least 2 pixel sections according to a preset sequence, wherein the absolute value of the difference value of the gray values of adjacent pixel points in the pixel sections is smaller than a preset first gray threshold, and the absolute value of the difference value of the gray values of a first pixel point and other pixel points in the pixel sections is smaller than a preset second gray threshold;
dividing the pixel segment into at least 2 sub-pixel segments according to a preset sequence, wherein the absolute value of the difference value of the parallax values of adjacent pixel points in the sub-pixel segments is smaller than a preset first parallax threshold, and the absolute value of the difference value of the parallax value of the first pixel point and other pixel points in the pixel segment is smaller than a preset second parallax threshold;
if the absolute value of the difference value of the parallax mean value of the sub-pixel segment and the pixel segment to which the sub-pixel segment belongs is larger than a preset third parallax threshold, determining that the sub-pixel segment is a sub-pixel segment with an abnormal parallax value;
and correcting the parallax value of the pixel point of the sub-pixel segment with the abnormal parallax value according to the parallax values of other sub-pixel segments, wherein the absolute value of the difference value between the adjacent parallax average value and the parallax average value of the pixel segment to which the adjacent parallax average value belongs is smaller than a preset third parallax threshold value.
In another aspect, the present invention provides a disparity value correction apparatus applied to a road scene, the apparatus including:
the disparity map calculation unit is used for calculating the disparity map of the reference map according to the reference map and the matching map of the same road scene;
the pixel segment dividing unit is used for dividing pixel points of each line in a road area of the reference image into at least 2 pixel segments according to a preset sequence, wherein the absolute value of the difference value of the gray values of adjacent pixel points in the pixel segments is smaller than a preset first gray threshold, and the absolute value of the difference value of the gray values of a first pixel point and other pixel points in the pixel segments is smaller than a preset second gray threshold;
the sub-pixel segment dividing unit is used for dividing the pixel segment into at least 2 sub-pixel segments according to a preset sequence, wherein the absolute value of the difference value of the parallax values of adjacent pixel points in the sub-pixel segments is smaller than a preset first parallax threshold, and the absolute value of the difference value of the parallax value of a first pixel point and the parallax value of other pixel points in the pixel segment is smaller than a preset second parallax threshold;
the abnormal sub-pixel segment determining unit is used for determining the sub-pixel segment as a parallax value abnormal sub-pixel segment if the absolute value of the difference value of the parallax mean values of the sub-pixel segment and the pixel segment to which the sub-pixel segment belongs is larger than a preset third parallax threshold;
and the parallax value correction unit is used for correcting the parallax value of the pixel point of the sub-pixel segment with the abnormal parallax value according to the parallax values of other sub-pixel segments, wherein the absolute value of the difference value between the adjacent parallax average value and the parallax average value of the pixel segment to which the adjacent parallax average value belongs is smaller than a preset third parallax threshold value.
As can be seen from the above description, the present invention first divides pixel segments based on the pixel points with close gray values in each line in the road area, then divides sub-pixel segments based on the pixel points with close parallax values in the pixel segments, when the sub-pixel segments with larger deviation between the parallax mean value and the parallax mean value of the pixel segments exist in the pixel segments, it indicates that the parallax fluctuation exists in the pixel segments with basically unchanged gray values (the probability that the pixel points with close gray values are located on the same object is higher, and the parallax values of the pixel points on the same object should be basically the same), determines the sub-pixel segments with larger deviation between the parallax mean value and the parallax mean value of the pixel segments as abnormal parallax values, corrects the parallax values of the sub-pixel segments with larger deviation between the parallax mean value and the parallax mean value of the pixel segments to which the adjacent parallax mean value is smaller than the parallax mean value of the pixel segments to which the adjacent parallax mean value is abnormal, that is to correct the parallax values of the sub-pixel segments with normal parallax values adjacent to the abnormal parallax values The parallax value of the pixel point of the abnormal sub-pixel segment is used for improving the accuracy of the parallax value of the road area, and the corrected parallax image is applied to auxiliary driving, so that the probability of false alarm can be effectively reduced.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings needed to be used in the description of the embodiments will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.
Fig. 1 is a flowchart illustrating a method for correcting a disparity value applied to a road scene according to an embodiment of the present invention;
FIG. 2A is a reference diagram of a road scene illustrating an embodiment of the present invention;
FIG. 2B is a matching graph of the same road scene as shown in the embodiment of the present invention;
FIG. 2C is a disparity map of the same road scene obtained by the SGM algorithm in the prior art;
FIG. 2D is a modified disparity map for the same road scene, according to an embodiment of the present invention;
FIG. 3 is a schematic diagram illustrating pixel segment division according to an embodiment of the present invention;
FIG. 4 is a schematic diagram of partitioning sub-pixel segments within a pixel segment according to an embodiment of the present invention;
FIG. 5 is a diagram illustrating a sub-pixel segment division according to an embodiment of the present invention;
fig. 6 is a schematic structural diagram of a binocular stereo platform device according to an embodiment of the present invention;
fig. 7 is a schematic structural diagram of a disparity value correction device applied to a road scene according to an embodiment of the present invention.
Detailed Description
Reference will now be made in detail to the exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, like numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present invention. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the invention, as detailed in the appended claims.
The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used in this specification and the appended claims, the singular forms "a", "an", and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It should also be understood that the term "and/or" as used herein refers to and encompasses any and all possible combinations of one or more of the associated listed items.
It is to be understood that although the terms first, second, third, etc. may be used herein to describe various information, these information should not be limited to these terms. These terms are only used to distinguish one type of information from another. For example, first information may also be referred to as second information, and similarly, second information may also be referred to as first information, without departing from the scope of the present invention. The word "if" as used herein may be interpreted as "at … …" or "when … …" or "in response to a determination", depending on the context.
The invention provides a parallax value correction method applied to a road scene, which is characterized in that pixel sections are divided based on pixel points with approximate gray values in each line in a road area, sub-pixel sections are divided based on the pixel points with the approximate parallax values in the pixel sections, the sub-pixel sections with larger parallax mean deviation and the parallax mean deviation of the pixel sections to which the sub-pixel sections belong are determined as parallax value abnormal sub-pixel sections, and the parallax values of the sub-pixel sections with the smaller parallax mean deviation and the adjacent parallax mean deviation and the parallax mean deviation of the pixel sections to which the sub-pixel sections belong are corrected to improve the accuracy of the parallax values in the road area.
Referring to fig. 1, a flowchart of an embodiment of a method for correcting a disparity value in a road scene according to the present invention is shown, and the embodiment describes a process for correcting a disparity value in a road region.
Step 101, calculating a disparity map of a reference map according to the reference map and a matching map of the same road scene.
In the auxiliary driving application, two vehicle-mounted depth cameras (also called binocular cameras) are used for shooting along a current driving road to obtain a depth map of a current road scene, the depth map comprises a left depth map and a right depth map, epipolar correction is carried out on the left depth map and the right depth map to obtain two depth maps which are parallel and equipotential, and matched pixel pairs are located on the same scanning line.
Taking a Semi-Global Matching (SGM) algorithm as an example, the SGM algorithm calculates a disparity value of each pixel in a reference map (one of the left depth map and the right depth map is the reference map, and the other is the Matching map) by using a pixel Matching result between the corrected left depth map and the right depth map, optimizes the disparity value by using an energy function, and finally determines an optimal disparity of each pixel in the reference map, so as to obtain the disparity map of the reference map.
Referring to fig. 2A, which is a reference diagram of a road scene according to an embodiment of the present invention, fig. 2B is a matching diagram of the same road scene, fig. 2C is a disparity diagram of the same road scene obtained by using the SGM algorithm, a disparity fluctuation exists in a frame region of a far road surface in the disparity diagram shown in fig. 2C, and it can be seen from the reference diagram shown in fig. 2A that there is no great texture change on the far road surface, that is, there is no abnormal situation, and the road surface is flat, and if the disparity diagram shown in fig. 2C is directly used for road surface detection, a false alarm is generated, which affects driving smoothness.
Step 102, dividing the pixel points of each line in the road area of the reference map into at least 2 pixel segments according to a preset sequence.
First, the road area is identified from the reference map, and the identification process can be implemented by using the prior art, which is not described herein again.
The present invention is not limited in this regard, and may be performed in a predetermined order, such as, for example, a left-to-right order, or a right-to-left order. For example, according to a left-to-right sequence, pixel segments are divided for each row in a road region in a reference image based on gray values of pixel points, an absolute value of a difference value of gray values of adjacent pixel points in the pixel segments is smaller than a preset first gray threshold, and absolute values of difference values of gray values of a first pixel point and gray values of other pixel points in the pixel segments are smaller than a preset second gray threshold, that is, a plurality of continuous pixel points with very close gray values are divided into one pixel segment.
Referring to fig. 3, a schematic diagram of dividing a pixel segment according to the embodiment of the present invention is shown, for example, sequentially scanning pixel points from left to right, and scanning a next pixel point P3 when an absolute value of a difference between a gray value of a pixel point P2 and a gray value of an adjacent pixel point P1 (the pixel point P1 is also a first pixel point of a current pixel segment) is smaller than a first gray threshold and is smaller than a second gray threshold; when the absolute value of the difference between the gray value of the pixel point P3 and the gray value of the adjacent pixel point P2 is smaller than a first gray threshold, and the absolute value of the difference between the gray value of the pixel point P3 and the gray value of the first pixel point P1 is smaller than a second gray threshold, scanning the next pixel point P4; when the absolute value of the difference between the gray value of the pixel point P4 and the gray value of the adjacent pixel point P3 is smaller than a first gray threshold, and the absolute value of the difference between the gray value of the pixel point P4 and the gray value of the first pixel point P1 is smaller than a second gray threshold, scanning the next pixel point P5; by analogy, until a pixel point which does not satisfy the gray scale condition is scanned, for example, if the absolute value of the difference between the gray scale value of the pixel point P17 and the gray scale value of the adjacent pixel point P16 is greater than a first gray scale threshold, the continuous pixel points from P1 to P16 are divided into one pixel segment, which is marked as H1, and then, the next pixel segment is continuously divided by taking P17 as the first pixel point of the next pixel segment.
It should be noted that, in order to avoid excessively long pixel segment partitions, the embodiment of the present invention limits the number of pixel points included in a pixel segment, and the preset number threshold, that is, the number of pixel points in a single pixel segment is smaller than the preset number threshold. Still taking fig. 3 as an example, if the preset number threshold is 17, even if the pixel point P17 satisfies the gray scale condition (i.e. the absolute value of the difference between the gray scale value of the pixel point P17 and the gray scale value of the adjacent pixel point P16 is smaller than the first gray scale threshold, and the absolute value of the difference between the gray scale value of the pixel point P17 and the gray scale value of the first pixel point P1 is smaller than the second gray scale threshold), the P17 can only be divided into the next pixel segment H2 due to the limitation of the number of pixels allowed to be included in a single pixel segment.
Step 103, dividing the pixel segment into at least 2 sub-pixel segments according to a preset sequence.
After the division of the pixel segments is completed through step 102, the sub-pixel segments are divided for each pixel segment based on the parallax values of the pixel points according to a preset sequence, for example, a left-to-right sequence, the absolute value of the difference value of the parallax values of the adjacent pixel points in the sub-pixel segment is smaller than a preset first parallax threshold, and the absolute value of the difference value of the parallax values of the first pixel point and the other pixel points in the sub-pixel segment is smaller than a preset second parallax threshold, that is, a plurality of continuous pixel points with similar parallax values are divided into one sub-pixel segment.
Referring to fig. 4, a schematic diagram of sub-pixel segment division according to an embodiment of the present invention is shown. Scanning the pixel segment H1 sequentially from left to right, and scanning a next pixel P3 when the absolute value of the difference between the parallax value of the pixel P2 and the parallax value of the adjacent pixel P1 (which is also the first pixel of the current sub-pixel segment) is smaller than the first parallax threshold and the second parallax threshold; when the absolute value of the difference between the parallax value of the pixel point P3 and the parallax value of the adjacent pixel point P2 is smaller than a first parallax threshold, and the absolute value of the difference between the parallax value of the pixel point P3 and the parallax value of the first pixel point P1 is smaller than a second parallax threshold, scanning the next pixel point P4; when the absolute value of the difference between the parallax value of the pixel point P4 and the parallax value of the adjacent pixel point P3 is smaller than a first parallax threshold, and the absolute value of the difference between the parallax value of the pixel point P4 and the parallax value of the pixel point P1 is smaller than a second parallax threshold, scanning the next pixel point P5; by analogy, until a pixel point which does not meet the parallax condition is scanned, for example, if the absolute value of the difference between the parallax value of the pixel point P6 and the parallax value of the adjacent pixel point P5 is greater than a first parallax threshold, it is determined that the pixel points P1 to P5 are a sub-pixel segment, and the sub-pixel segment is marked as S1; p6 is the first pixel point of the next sub-pixel segment S2, and is divided in the same manner as S1, as shown in fig. 4, and finally, the pixel segment H1 is divided into three sub-pixel segments S1, S2, and S3. Fig. 4 is an exemplary illustration only, and the number of sub-pixel segments in a pixel segment is not limited in the embodiments of the present invention.
And 104, if the absolute value of the difference value of the parallax mean value of the sub-pixel segment and the pixel segment to which the sub-pixel segment belongs is greater than a preset third parallax threshold, determining that the sub-pixel segment is a sub-pixel segment with an abnormal parallax value.
In the step, whether the parallax value of the sub-pixel segment is abnormal is determined based on the parallax average value of the sub-pixel segment (the average value of the parallax values of the pixels in the sub-pixel segment) and the parallax average value of the pixel segment to which the sub-pixel segment belongs (the average value of the parallax values of the pixels in the pixel segment), and the sub-pixel segment with the abnormal parallax value is referred to as the parallax value abnormal sub-pixel segment.
Specifically, if the absolute value of the difference between the disparity averages of the sub-pixel segment and the pixel segment to which the sub-pixel segment belongs is greater than the preset third disparity threshold, the sub-pixel segment is determined to be the disparity value abnormal sub-pixel segment, because the pixel segment is divided based on the pixel points with the similar gray values in the step 102, the disparity values of the pixel points in the pixel segment should be substantially the same under normal conditions, and therefore, when it is determined that the disparity average is greatly deviated from the disparity average of the sub-pixel segment, the sub-pixel segment is the disparity value abnormal sub-pixel segment.
And 105, correcting the parallax value of the pixel point of the sub-pixel segment with the abnormal parallax value according to the parallax values of other sub-pixel segments, wherein the absolute value of the difference value between the adjacent parallax average value and the parallax average value of the pixel segment to which the adjacent parallax average value belongs is smaller than a preset third parallax threshold value.
In this step, the other sub-pixel segments whose absolute value of the difference between the parallax average value and the parallax average value of the pixel segment to which the parallax average value belongs is smaller than the preset third parallax threshold value refer to the sub-pixel segments whose parallax average value has smaller deviation from the parallax average value of the pixel segment to which the parallax average value belongs, that is, the sub-pixel segments which do not cause large parallax fluctuation in the pixel segment to which the parallax average value belongs are determined as the sub-pixel segments whose parallax values are normal, and in this step, the parallax values of the pixel points of the sub-pixel segments whose parallax values are abnormal are corrected by using the parallax values of the sub-pixel segments whose parallax values are normal adjacent to the sub-pixel segments whose parallax values are abnormal. The specific modification mode is as follows:
the first embodiment: and when the absolute value of the difference value between the parallax average value of other sub-pixel sections adjacent to the parallax value abnormal sub-pixel section and the parallax average value of the pixel section to which the other sub-pixel sections belong is smaller than a preset third parallax threshold value, and the other sub-pixel sections and the parallax value abnormal sub-pixel section belong to the same pixel section, correcting the parallax value of the pixel point of the parallax value abnormal sub-pixel section according to the parallax values of the other sub-pixel sections. That is, when the sub-pixel segment adjacent to the sub-pixel segment with abnormal parallax value is the sub-pixel segment with normal parallax value, and the sub-pixel segment with normal parallax value and the sub-pixel segment with abnormal parallax value belong to the same pixel segment, the parallax values should be substantially the same because the gray values of the pixels in the same pixel segment are close, so that the parallax value of the pixel of the sub-pixel segment with abnormal parallax value can be corrected by using the parallax value of the sub-pixel segment with normal parallax value adjacent to the sub-pixel segment with abnormal parallax value.
Referring to fig. 5, if S2 is the sub-pixel segment with abnormal parallax value, S1 is the sub-pixel segment with normal parallax value, and S3 is the sub-pixel segment with abnormal parallax value, the parallax value of the pixel point in S2 is corrected, and since S1 is the sub-pixel segment with normal parallax value, which is adjacent to S2 and belongs to the same pixel segment H1, the parallax value of the pixel point in S1 can be used to correct the parallax value of the pixel point in S2. Similarly, if S3 is also the sub-pixel segment with normal parallax value, S1, S2 and S3 all belong to the pixel segment H1, so the parallax value of the pixel point in S1 and S3 can be used to correct the parallax value of the pixel segment in S2.
The second embodiment: and when the absolute value of the difference value between the parallax average value of other sub-pixel segments adjacent to the parallax value abnormal sub-pixel segment and the parallax average value of the pixel segment to which the sub-pixel segment belongs is smaller than a preset third parallax threshold, the other sub-pixel segments and the parallax value abnormal sub-pixel segment belong to 2 adjacent pixel segments respectively, the absolute value of the difference value between the gray values of the adjacent pixel points of the other sub-pixel segments and the parallax value abnormal sub-pixel segment is smaller than the first gray threshold, and the absolute value of the difference value between the gray values of the first pixel point or the last pixel point of the other sub-pixel segments and the gray value of the first pixel point of the pixel segment to which the parallax value abnormal sub-pixel segment belongs is smaller than the second gray threshold, correcting the parallax value of the pixel point of the parallax value abnormal sub-pixel segment according to the parallax values of the other sub-pixel segments. That is, the sub-pixel segment with normal parallax value adjacent to the sub-pixel segment with abnormal parallax value and the sub-pixel segment with abnormal parallax value belong to the adjacent pixel segment, and the gray values of the sub-pixel segments with abnormal parallax value are close to the gray values of the sub-pixel segments with normal parallax value, it is possible to divide the pixels with close gray values into the adjacent pixel segments respectively due to the limitation of the division length of the pixel segments (the number threshold of the pixels in the pixel segments is reached), and therefore, even if the sub-pixel segment with normal parallax value adjacent to the sub-pixel segment with abnormal parallax value and the sub-pixel segment with abnormal parallax value do not belong to the same pixel segment, the parallax value of the pixel point of the sub-pixel segment with abnormal parallax value can be corrected by utilizing the parallax value of the sub-pixel segment with normal parallax value adjacent to the sub-pixel segment with abnormal parallax value, wherein the gray value of the sub-pixel segment is close to the gray value of the sub-pixel segment with abnormal parallax value.
Referring to fig. 5, if S3 is the sub-pixel segment with abnormal parallax value, S2 is the sub-pixel segment with abnormal parallax value, and S4 is the sub-pixel segment with normal parallax value, the parallax value of the pixel in S3 is corrected. As shown in fig. 5, S3 and S4 are adjacent to each other and belong to adjacent pixel segments H1 and H2, if the absolute value of the difference between the gray value of the pixel P17 in S4 adjacent to the pixel P16 in S3 and the gray value of P16 is smaller than the first gray threshold, and the absolute value of the difference between the gray value of the pixel P17 and the gray value of the first pixel P1 in the pixel segment H1 belonging to S3 is smaller than the second gray threshold, it is indicated that the gray values of the pixels in P17 and H1 are close to each other, but the pixel segments are artificially divided into the pixel segments H2 due to the limitation of the maximum allowable number of the pixels in the pixel segments, so that the sub-pixel segment S4 with a normal parallax value in the pixel segment H2 should have a parallax value close to S3, and therefore, the parallax value of the pixel in S3 can be corrected by using the parallax value of the pixel in S4. Of course, if S2 is also the sub-pixel segment with normal parallax value, the parallax value of the pixel in S3 can be corrected by combining the parallax values of the pixels in S2 and S4.
Third embodiment: and when the absolute value of the difference value between the parallax average value of the front sub-pixel segment and the parallax average value of the other sub-pixel segment adjacent to the parallax value abnormal sub-pixel segment and the parallax average value of the pixel segment to which the parallax average value belongs is smaller than a preset third parallax threshold value, and the absolute value of the difference value between the average value of the gray values of the two other sub-pixel segments before and after the parallax value abnormal sub-pixel segment and the average value of the gray values of the parallax value abnormal sub-pixel segment is smaller than the third gray threshold value, correcting the parallax value of the pixel point of the parallax value abnormal sub-pixel segment according to the parallax value of the front sub-pixel segment and the rear sub-pixel segment adjacent to the parallax value abnormal sub-pixel segment. That is, when the front and rear sub-pixel segments adjacent to the sub-pixel segment with the abnormal parallax value are sub-pixel segments with the normal parallax value, and the average value of the gray values of the front and rear sub-pixel segments with the normal parallax value is close to the average value of the gray values of the sub-pixel segments with the abnormal parallax value, the parallax value of the pixel point of the sub-pixel segment with the abnormal parallax value is corrected by using the parallax value of the front and rear sub-pixel segments with the normal parallax value.
Referring to fig. 5, if S3 is the sub-pixel segment with abnormal parallax value, S2 is the sub-pixel segment with normal parallax value, and S4 is the sub-pixel segment with normal parallax value, the parallax value of the pixel point in S3 is corrected. In this embodiment, it is not necessary to determine whether S2, S4 belong to the same pixel segment as S3 or belong to adjacent pixel segments, but it is determined whether S2, S4 can be used to correct the parallax value of the pixel in S3 based on the gray values of S2, S3, S4. Specifically, the average of the gray values of the pixels in S2 and S4 is calculated, and if the absolute value of the difference between the average of the gray values and the average of the gray values of the pixels in S3 is smaller than the third gray threshold, the parallax value of the pixel in S3 is corrected by using the parallax values of the pixels in S2 and S4.
Fourth embodiment: and when the absolute value of the difference value between the parallax average value of the front sub-pixel segment and the parallax average value of the other sub-pixel segments adjacent to the parallax value abnormal sub-pixel segment and the parallax average value of the pixel segment to which the parallax average value belongs is smaller than a preset third parallax threshold, and the absolute value of the difference value between the average value of the gray values of the other sub-pixel segments before or after the parallax value abnormal sub-pixel segment and the average value of the gray values of the parallax value abnormal sub-pixel segment is smaller than the third gray threshold, correcting the parallax value of the pixel point of the parallax value abnormal sub-pixel segment according to the parallax value of the other sub-pixel segments of which the absolute value of the difference value with the average value of the gray values of the parallax value abnormal sub-pixel segment is smaller than the. That is, when the front and rear sub-pixel segments adjacent to the sub-pixel segment with abnormal parallax value are all sub-pixel segments with normal parallax value, and the average of the gray values of the sub-pixel segments with normal parallax value before or after the sub-pixel segment with abnormal parallax value is close to the average of the gray values of the sub-pixel segments with abnormal parallax value, the parallax value of the pixel point of the sub-pixel segment with abnormal parallax value is corrected by using the parallax value of the sub-pixel segment with normal parallax value close to the average of the gray values of the sub-pixel segments with abnormal parallax value.
Referring to fig. 5, if S3 is the sub-pixel segment with abnormal parallax value, S2 is the sub-pixel segment with normal parallax value, and S4 is the sub-pixel segment with normal parallax value, the parallax value of the pixel point in S3 is corrected. And respectively calculating the average value of the gray values of the pixels in the S2 and the average value of the gray values of the pixels in the S4, and if the absolute value of the difference between the average value of the gray values of the pixels in the S2 and the average value of the gray values of the pixels in the S3 is greater than a third gray threshold value, and the absolute value of the difference between the average value of the gray values of the pixels in the S4 and the average value of the gray values of the pixels in the S3 is smaller than the third gray threshold value, correcting the parallax value of the pixels in the S3 by using the parallax value of the pixels in the.
In the above embodiments, the parallax value of the pixel point of the parallax value abnormal sub-pixel segment is corrected based on the parallax value of the pixel point of the parallax value normal sub-pixel segment close to the gray value adjacent to the parallax value abnormal sub-pixel segment, and during correction, the parallax value of the parallax value abnormal sub-pixel segment can be smoothly processed based on the change rule of the parallax value of the adjacent parallax value normal sub-pixel segment.
It should be added that, when the absolute values of the differences between the parallax mean values of the front and rear sub-pixel segments adjacent to the sub-pixel segment with the abnormal parallax value and the parallax mean value of the pixel segment to which the sub-pixel segment belongs are both greater than the preset third parallax threshold, the pixel points in which the difference between the parallax mean value of the sub-pixel segment with the abnormal parallax value and the parallax mean value of the sub-pixel segment with the abnormal parallax value is greater than the preset fourth parallax threshold are deleted. That is, when the front and rear sub-pixel segments adjacent to the sub-pixel segment with abnormal parallax value are sub-pixel segments with abnormal parallax value, the parallax value of the pixel point in the adjacent sub-pixel segment cannot be used to correct the parallax value of the pixel point in the sub-pixel segment with abnormal parallax value.
Through the processing process, the correction of the parallax value of the pixel points in the parallax value abnormal sub-pixel segment of each line in the road area can be completed, and the fluctuation of the parallax value in the horizontal direction is restrained. After the correction of the parallax value of the pixel point in the horizontal direction (row direction) is completed, the parallax value of the pixel point in the vertical direction (column direction) in the road area may be further smoothed, for example, the pixel point in the vertical direction is smoothed by applying a nonlinear fitting method in a segmented manner.
From the above description, it can be seen that the present invention divides pixel segments based on the pixel points with close gray values in each row in the road area, divides sub-pixel segments based on the pixel points with close parallax values in the pixel segments, when the sub-pixel segments with larger deviation between the parallax mean value and the parallax mean value of the pixel segments exist in the pixel segments, the sub-pixel segments with larger deviation between the parallax mean value and the parallax mean value of the pixel segments are determined as the abnormal parallax value sub-pixel segments, the parallax values of the pixel points of the abnormal parallax value sub-pixel segments are corrected by using the parallax values of the sub-pixel segments with smaller deviation between the adjacent parallax mean values and the parallax mean values of the pixel segments to which the adjacent parallax mean values are normal, i.e. the parallax values of the sub-pixel segments with normal parallax values adjacent to the abnormal parallax value sub-pixel segments are used to correct the parallax values of the abnormal parallax values of the sub-pixel segments, so as to improve the accuracy of the parallax values of the road area, and apply the corrected parallax images to assistant driving, the probability of false alarm can be effectively reduced.
Referring to fig. 2A, a depth map of a road scene shot by one vehicle-mounted depth camera, and fig. 2B, a depth map of the same road scene shot by another vehicle-mounted depth camera, where fig. 2A is a reference map, and fig. 2B is a matching map, and when a disparity map of the road scene is calculated based on the reference map 2A and the matching map 2B by using an SGM algorithm in the prior art, a disparity value of each pixel point in the reference map 2A is calculated by using matching results of the pixel points in the reference map 2A and the matching map 2B, and a final energy value of each pixel point is obtained by using energy function propagation, and an optimal disparity value of the pixel point is determined, so that the disparity map of the road scene shown in fig. 2C is obtained. Referring to fig. 2C, due to the different shooting angles or different exposures of the two vehicle-mounted depth cameras, parallax fluctuation occurs in the frame region shown in fig. 2C, and it can be seen from the reference map 2A and the matching map 2B that the far road region is flat and has no abnormal condition, and if the road detection is performed based on the parallax map shown in fig. 2C, false alarm occurs, which affects driving smoothness.
However, referring to fig. 2D, the disparity map shown in fig. 2D is a disparity map obtained by performing disparity value correction on a road area by using the method for correcting disparity values according to the embodiment of the present invention. According to the embodiment of the invention, the pixel segments are divided based on the pixel points with the approximate gray values, then the pixel segments are divided into the sub-pixel segments based on the pixel points with the approximate parallax values, and the sub-pixel segments with the abnormal parallax values are corrected by using the parallax values of the sub-pixel segments with the normal parallax values, which are approximate to the gray values of the sub-pixel segments, so that the regions with the approximate gray values and fluctuating parallax values in the road-dropping region are filtered, the accuracy of the corrected parallax values is improved, and the false alarm probability of road detection in auxiliary driving is reduced.
Corresponding to the foregoing embodiments of the method for correcting a disparity value in a road scene, the present invention also provides embodiments of a device for correcting a disparity value in a road scene.
The embodiment of the parallax value correction device applied to the road scene can be applied to binocular stereo vision platform equipment. The device embodiments may be implemented by software, or by hardware, or by a combination of hardware and software. A software implementation is taken as an example, and a logical means is formed by a processor of the device in which it is located running corresponding computer program instructions in a memory. In terms of hardware, as shown in fig. 6, a hardware structure diagram of a device where the disparity value correction apparatus is applied in a road scene according to the present invention is shown, except for the processor and the nonvolatile memory shown in fig. 6, the device where the apparatus is located in the embodiment may also include other hardware according to the actual function of the device, which is not described again.
Fig. 7 is a schematic structural diagram of a disparity value correction device applied to a road scene according to an embodiment of the present invention. The parallax value correction device applied to the road scene comprises: a disparity map calculation unit 701, a pixel segment division unit 702, a sub-pixel segment division unit 703, an abnormal sub-pixel segment determination unit 704, and a disparity value correction unit 705, wherein:
a disparity map calculation unit 701 configured to calculate a disparity map of a reference map based on a reference map and a matching map of the same road scene;
a pixel segment dividing unit 702, configured to divide, according to a preset sequence, pixel points in each row in a road region of the reference map into at least 2 pixel segments, where an absolute value of a difference between gray values of adjacent pixel points in each pixel segment is smaller than a preset first gray threshold, and absolute values of differences between gray values of a first pixel point and other pixel points in each pixel segment are smaller than a preset second gray threshold;
a sub-pixel segment dividing unit 703, configured to divide the pixel segment into at least 2 sub-pixel segments according to a preset sequence, where an absolute value of a difference between disparity values of adjacent pixels in the sub-pixel segment is smaller than a preset first disparity threshold, and absolute values of differences between disparity values of a first pixel and other pixels in the sub-pixel segment are both smaller than a preset second disparity threshold;
an abnormal sub-pixel segment determining unit 704, configured to determine that the sub-pixel segment is a sub-pixel segment with an abnormal disparity value if an absolute value of a difference between disparity averages of the sub-pixel segment and a pixel segment to which the sub-pixel segment belongs is greater than a preset third disparity threshold;
the parallax value correcting unit 705 is configured to correct the parallax value of the pixel point of the sub-pixel segment with the abnormal parallax value according to the parallax values of the other sub-pixel segments, where an absolute value of a difference value between the adjacent parallax average value and the parallax average value of the pixel segment to which the adjacent parallax average value belongs is smaller than a preset third parallax threshold.
Further, the air conditioner is provided with a fan,
the disparity value correcting unit 705 is specifically configured to correct, when an absolute value of a difference between a disparity average value of another sub-pixel segment adjacent to the disparity value abnormal sub-pixel segment and a disparity average value of a pixel segment to which the sub-pixel segment belongs is smaller than a preset third disparity threshold, and the other sub-pixel segment and the disparity value abnormal sub-pixel segment belong to the same pixel segment, a disparity value of a pixel point of the disparity value abnormal sub-pixel segment according to the disparity values of the other sub-pixel segments.
Further, the air conditioner is provided with a fan,
the disparity value modification unit 705 is specifically configured to, when the absolute value of the difference between the disparity average value of the other sub-pixel segment adjacent to the disparity value abnormal sub-pixel segment and the disparity average value of the pixel segment to which the other sub-pixel segment belongs is smaller than a preset third disparity threshold, and the other sub-pixel segments and the sub-pixel segment with abnormal parallax value belong to 2 adjacent pixel segments respectively, the absolute value of the difference value of the gray values of the adjacent pixel points of the other sub-pixel segments and the parallax value abnormal sub-pixel segment is smaller than the first gray threshold value, and the absolute value of the difference value of the gray values of the first pixel point or the last pixel point of the other sub-pixel segments and the first pixel point of the pixel segment of the sub-pixel segment with the abnormal parallax value is smaller than the second gray threshold, and correcting the parallax value of the pixel point of the sub-pixel segment with the abnormal parallax value according to the parallax values of the other sub-pixel segments.
Further, the air conditioner is provided with a fan,
the disparity value correcting unit 705 is specifically configured to correct the disparity value of the pixel point of the disparity value abnormal sub-pixel segment according to the disparity value of the two sub-pixel segments before and after the disparity value abnormal sub-pixel segment, where the absolute value of the difference between the disparity average value of the two sub-pixel segments before and after the disparity value abnormal sub-pixel segment and the disparity average value of the gray value of the disparity value abnormal sub-pixel segment is smaller than a preset third disparity threshold, and the absolute value of the difference between the average value of the gray values of the two sub-pixel segments before and after the disparity value abnormal sub-pixel segment and the average value of the gray value of the disparity value abnormal sub-pixel segment is smaller than the third gray threshold.
Or, the disparity value correcting unit 705 is specifically configured to correct the disparity value of the pixel point of the disparity value abnormal sub-pixel segment according to the disparity value of the other sub-pixel segment whose absolute value of the difference from the average value of the gray value of the disparity value abnormal sub-pixel segment is smaller than the third gray threshold when the absolute value of the difference between the disparity average value of the front and back two adjacent sub-pixel segments adjacent to the disparity value abnormal sub-pixel segment and the disparity average value of the pixel segment to which the sub-pixel segment belongs is smaller than the preset third disparity threshold, and the absolute value of the difference between the average value of the gray value of the other sub-pixel segment before or after the disparity value abnormal sub-pixel segment and the average value of the gray value of the disparity value abnormal sub-pixel segment is smaller than the third gray threshold.
Or, the parallax value correcting unit 705 is further configured to delete, when the absolute values of the differences between the parallax mean values of the two previous and next sub-pixel segments adjacent to the sub-pixel segment with the abnormal parallax value and the parallax mean value of the pixel segment to which the sub-pixel segment belongs are both greater than a preset third parallax threshold, the pixel point in which the difference between the parallax value in the sub-pixel segment with the abnormal parallax value and the parallax mean value of the sub-pixel segment with the abnormal parallax value is greater than a preset fourth parallax threshold.
The implementation process of the functions and actions of each unit in the above device is specifically described in the implementation process of the corresponding step in the above method, and is not described herein again.
For the device embodiments, since they substantially correspond to the method embodiments, reference may be made to the partial description of the method embodiments for relevant points. The above-described embodiments of the apparatus are merely illustrative, and the units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the scheme of the invention. One of ordinary skill in the art can understand and implement it without inventive effort.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like made within the spirit and principle of the present invention should be included in the scope of the present invention.
Claims (10)
1. A method for correcting a parallax value applied to a road scene, the method comprising:
calculating a disparity map of the reference map according to the reference map and the matching map of the same road scene;
dividing pixel points in each row in a road area of the reference image into at least 2 pixel sections according to a preset sequence, wherein the absolute value of the difference value of the gray values of adjacent pixel points in the pixel sections is smaller than a preset first gray threshold, and the absolute value of the difference value of the gray values of a first pixel point and other pixel points in the pixel sections is smaller than a preset second gray threshold;
dividing the pixel segment into at least 2 sub-pixel segments according to a preset sequence, wherein the absolute value of the difference value of the parallax values of adjacent pixels in the sub-pixel segments is smaller than a preset first parallax threshold, and the absolute value of the difference value of the parallax value of a first pixel and other pixels in the sub-pixel segments is smaller than a preset second parallax threshold;
if the absolute value of the difference value between the parallax average value of the sub-pixel segment and the parallax average value of the pixel segment to which the sub-pixel segment belongs is larger than a preset third parallax threshold, determining that the sub-pixel segment is a sub-pixel segment with an abnormal parallax value;
and correcting the parallax value of the pixel point of the sub-pixel segment with the abnormal parallax value according to the parallax values of other sub-pixel segments, wherein the absolute value of the difference value between the adjacent parallax average value and the parallax average value of the pixel segment to which the adjacent parallax average value belongs is smaller than a preset third parallax threshold value.
2. The method according to claim 1, wherein the correcting the disparity value of the pixel point of the sub-pixel segment with the abnormal disparity value according to the disparity value of the other sub-pixel segment whose absolute value of the difference between the adjacent disparity mean value and the disparity mean value of the pixel segment to which the adjacent disparity mean value belongs is smaller than a preset third disparity threshold specifically comprises:
and when the absolute value of the difference value between the parallax average value of other sub-pixel sections adjacent to the parallax value abnormal sub-pixel section and the parallax average value of the pixel section to which the other sub-pixel sections belong is smaller than a preset third parallax threshold value, and the other sub-pixel sections and the parallax value abnormal sub-pixel section belong to the same pixel section, correcting the parallax value of the pixel point of the parallax value abnormal sub-pixel section according to the parallax values of the other sub-pixel sections.
3. The method according to claim 1, wherein the correcting the disparity value of the pixel point of the sub-pixel segment with the abnormal disparity value according to the disparity value of the other sub-pixel segment whose absolute value of the difference between the adjacent disparity mean value and the disparity mean value of the pixel segment to which the adjacent disparity mean value belongs is smaller than a preset third disparity threshold specifically comprises:
when the absolute value of the difference value between the parallax average value of other sub-pixel segments adjacent to the parallax value abnormal sub-pixel segment and the parallax average value of the pixel segment to which the sub-pixel segment belongs is smaller than a preset third parallax threshold, the other sub-pixel segments and the parallax value abnormal sub-pixel segment respectively belong to 2 adjacent pixel segments, the absolute value of the difference value between the gray values of the adjacent pixel points of the other sub-pixel segments and the parallax value abnormal sub-pixel segment is smaller than the first gray threshold, and the absolute value of the difference value between the gray value of the first pixel point or the gray value of the last pixel point of the other sub-pixel segments and the gray value of the first pixel point of the pixel segment to which the parallax value abnormal sub-pixel segment belongs is smaller than the second gray threshold, the parallax value of the pixel point of the parallax value abnormal sub-pixel segment is corrected according to the parallax values of the other sub-pixel segments.
4. The method according to claim 1, wherein the correcting the disparity value of the pixel point of the sub-pixel segment with the abnormal disparity value according to the disparity value of the other sub-pixel segment whose absolute value of the difference between the adjacent disparity mean value and the disparity mean value of the pixel segment to which the adjacent disparity mean value belongs is smaller than a preset third disparity threshold specifically comprises:
and when the absolute value of the difference value between the parallax average value of the front sub-pixel segment and the parallax average value of the other sub-pixel segment adjacent to the parallax value abnormal sub-pixel segment and the parallax average value of the pixel segment to which the parallax value abnormal sub-pixel segment belongs is smaller than a preset third parallax threshold, and the absolute value of the difference value between the average value of the gray values of the two other sub-pixel segments before and after the parallax value abnormal sub-pixel segment and the average value of the gray values of the parallax value abnormal sub-pixel segment is smaller than a third gray threshold, correcting the parallax value of the pixel point of the parallax value abnormal sub-pixel segment according to the parallax value of the front sub-pixel segment and the rear sub-pixel segment adjacent to the parallax value abnormal sub-pixel segment.
5. The method according to claim 1, wherein the correcting the disparity value of the pixel point of the sub-pixel segment with the abnormal disparity value according to the disparity value of the other sub-pixel segment whose absolute value of the difference between the adjacent disparity mean value and the disparity mean value of the pixel segment to which the adjacent disparity mean value belongs is smaller than a preset third disparity threshold specifically comprises:
and when the absolute value of the difference value between the parallax average value of the front sub-pixel segment and the parallax average value of the other sub-pixel segment adjacent to the parallax value abnormal sub-pixel segment and the parallax average value of the pixel segment to which the front sub-pixel segment and the rear sub-pixel segment belong is smaller than a preset third parallax threshold, and the absolute value of the difference value between the average value of the gray values of the other sub-pixel segments before or after the parallax value abnormal sub-pixel segment and the average value of the gray values of the parallax value abnormal sub-pixel segment is smaller than the third gray threshold, correcting the parallax value of the pixel point of the parallax value abnormal sub-pixel segment according to the parallax value of the other sub-pixel segment of which the absolute value of the difference value with the average value of the gray values of the parallax value abnormal sub-pixel segment is.
6. The method according to any one of claims 1 to 5, wherein if the absolute value of the difference value between the disparity mean values of the sub-pixel segment and the pixel segment to which the sub-pixel segment belongs is greater than a preset third disparity threshold, determining that the sub-pixel segment is a disparity value abnormal sub-pixel segment, and then further comprising:
and when the absolute value of the difference value between the parallax mean value of the front sub-pixel segment and the parallax mean value of the other front sub-pixel segment and the rear sub-pixel segment adjacent to the parallax value abnormal sub-pixel segment and the parallax mean value of the pixel segment to which the parallax value abnormal sub-pixel segment belongs is larger than a preset third parallax threshold, deleting the pixel points of which the difference value between the parallax mean value in the parallax value abnormal sub-pixel segment and the parallax mean value of the parallax value abnormal sub-pixel segment is larger than a preset fourth parallax threshold.
7. A disparity value correction apparatus applied to a road scene, the apparatus comprising:
the disparity map calculation unit is used for calculating the disparity map of the reference map according to the reference map and the matching map of the same road scene;
the pixel segment dividing unit is used for dividing pixel points of each line in a road area of the reference image into at least 2 pixel segments according to a preset sequence, wherein the absolute value of the difference value of the gray values of adjacent pixel points in the pixel segments is smaller than a preset first gray threshold, and the absolute value of the difference value of the gray values of a first pixel point and other pixel points in the pixel segments is smaller than a preset second gray threshold;
the sub-pixel segment dividing unit is used for dividing the pixel segment into at least 2 sub-pixel segments according to a preset sequence, wherein the absolute value of the difference value of the parallax values of adjacent pixel points in the sub-pixel segments is smaller than a preset first parallax threshold, and the absolute value of the difference value of the parallax value of a first pixel point and the parallax value of other pixel points in the sub-pixel segments is smaller than a preset second parallax threshold;
the abnormal sub-pixel segment determining unit is used for determining the sub-pixel segment as a parallax value abnormal sub-pixel segment if the absolute value of the difference value between the parallax average value of the sub-pixel segment and the parallax average value of the pixel segment to which the sub-pixel segment belongs is larger than a preset third parallax threshold;
and the parallax value correction unit is used for correcting the parallax value of the pixel point of the sub-pixel segment with the abnormal parallax value according to the parallax values of other sub-pixel segments, wherein the absolute value of the difference value between the adjacent parallax average value and the parallax average value of the pixel segment to which the adjacent parallax average value belongs is smaller than a preset third parallax threshold value.
8. The apparatus of claim 7, wherein:
the parallax value correcting unit is specifically configured to correct the parallax value of the pixel point of the parallax value abnormal sub-pixel segment according to the parallax values of the other sub-pixel segments when the absolute value of the difference between the parallax average value of the other sub-pixel segments adjacent to the parallax value abnormal sub-pixel segment and the parallax average value of the pixel segment to which the other sub-pixel segments belong is smaller than a preset third parallax threshold and the other sub-pixel segments and the parallax value abnormal sub-pixel segment belong to the same pixel segment.
9. The apparatus of claim 7, wherein:
the parallax value correcting unit is specifically configured to, when the absolute value of the difference between the parallax mean value of the other sub-pixel segment adjacent to the parallax value abnormal sub-pixel segment and the parallax mean value of the pixel segment to which the other sub-pixel segment belongs is smaller than a preset third parallax threshold, and the other sub-pixel segments and the sub-pixel segment with abnormal parallax value belong to 2 adjacent pixel segments respectively, the absolute value of the difference value of the gray values of the adjacent pixel points of the other sub-pixel segments and the parallax value abnormal sub-pixel segment is smaller than the first gray threshold value, and the absolute value of the difference value of the gray values of the first pixel point or the last pixel point of the other sub-pixel segments and the first pixel point of the pixel segment of the sub-pixel segment with the abnormal parallax value is smaller than the second gray threshold, and correcting the parallax value of the pixel point of the sub-pixel segment with the abnormal parallax value according to the parallax values of the other sub-pixel segments.
10. The apparatus of claim 7, wherein:
the parallax value correcting unit is specifically configured to correct the parallax value of the pixel point of the parallax value abnormal sub-pixel segment according to the parallax values of the two sub-pixel segments before and after the parallax value abnormal sub-pixel segment, when the absolute value of the difference between the parallax average value of the two sub-pixel segments adjacent to the parallax value abnormal sub-pixel segment and the parallax average value of the pixel segment to which the parallax average value belongs is smaller than a preset third parallax threshold, and the absolute value of the difference between the average value of the gray values of the two sub-pixel segments before and after the parallax value abnormal sub-pixel segment and the average value of the gray values of the parallax value abnormal sub-pixel segment is smaller than a third gray threshold;
or the parallax value correcting unit is specifically configured to correct the parallax value of the pixel point of the parallax value abnormal sub-pixel segment according to the parallax value of the other sub-pixel segment whose absolute value of the difference from the average value of the gray value of the parallax value abnormal sub-pixel segment is smaller than a third gray threshold when the absolute value of the difference between the parallax average value of the two adjacent sub-pixel segments adjacent to the parallax value abnormal sub-pixel segment and the parallax average value of the pixel segment to which the two adjacent sub-pixel segments belong is smaller than a preset third parallax threshold, and the absolute value of the difference between the average value of the gray values of the other sub-pixel segments before or after the parallax value abnormal sub-pixel segment and the average value of the gray value of the parallax value abnormal sub-pixel segment is smaller than the third gray threshold;
or the parallax value correction unit is further configured to delete, when the absolute values of the differences between the parallax mean values of the two adjacent sub-pixel segments adjacent to the abnormal parallax value sub-pixel segment and the parallax mean values of the pixel segments to which the two adjacent sub-pixel segments belong are both greater than a preset third parallax threshold, the pixel points in the abnormal parallax value sub-pixel segment whose difference between the parallax mean value and the parallax mean value of the abnormal parallax value sub-pixel segment is greater than a preset fourth parallax threshold.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710761404.6A CN107564053B (en) | 2017-08-28 | 2017-08-28 | Parallax value correction method and device applied to road scene |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710761404.6A CN107564053B (en) | 2017-08-28 | 2017-08-28 | Parallax value correction method and device applied to road scene |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107564053A CN107564053A (en) | 2018-01-09 |
| CN107564053B true CN107564053B (en) | 2020-06-09 |
Family
ID=60977823
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710761404.6A Active CN107564053B (en) | 2017-08-28 | 2017-08-28 | Parallax value correction method and device applied to road scene |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107564053B (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104537668A (en) * | 2014-12-29 | 2015-04-22 | 浙江宇视科技有限公司 | Fast anaglyph calculating method and device |
| CN104915941A (en) * | 2014-03-11 | 2015-09-16 | 株式会社理光 | Method and apparatus for calculating parallax |
| CN105550665A (en) * | 2016-01-15 | 2016-05-04 | 北京理工大学 | Method for detecting pilotless automobile through area based on binocular vision |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7903849B2 (en) * | 2005-04-15 | 2011-03-08 | Kabushiki Kaisha Toshiba | Medical image processing apparatus |
| JP6554973B2 (en) * | 2015-07-29 | 2019-08-07 | ブラザー工業株式会社 | Image reading device |
-
2017
- 2017-08-28 CN CN201710761404.6A patent/CN107564053B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104915941A (en) * | 2014-03-11 | 2015-09-16 | 株式会社理光 | Method and apparatus for calculating parallax |
| CN104915941B (en) * | 2014-03-11 | 2017-08-04 | 株式会社理光 | The method and apparatus for calculating parallax |
| CN104537668A (en) * | 2014-12-29 | 2015-04-22 | 浙江宇视科技有限公司 | Fast anaglyph calculating method and device |
| CN105550665A (en) * | 2016-01-15 | 2016-05-04 | 北京理工大学 | Method for detecting pilotless automobile through area based on binocular vision |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107564053A (en) | 2018-01-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10097812B2 (en) | Stereo auto-calibration from structure-from-motion | |
| US8755592B2 (en) | Stereo matching system using dynamic programming and method thereof | |
| US9237326B2 (en) | Imaging system and method | |
| JP6291469B2 (en) | Obstacle detection device and obstacle detection method | |
| CN105069804B (en) | Threedimensional model scan rebuilding method based on smart mobile phone | |
| CN104915927B (en) | Anaglyph optimization method and device | |
| CN111612728B (en) | 3D point cloud densification method and device based on binocular RGB image | |
| US9769460B1 (en) | Conversion of monoscopic visual content to stereoscopic 3D | |
| US8989481B2 (en) | Stereo matching device and method for determining concave block and convex block | |
| CN106651897B (en) | Parallax correction method based on super-pixel segmentation | |
| US9582859B2 (en) | System and a method for depth-image-based rendering | |
| CN107392898B (en) | Method and device for calculating pixel point parallax value applied to binocular stereo vision | |
| CN108109148A (en) | Image solid distribution method, mobile terminal | |
| CN113744315A (en) | Semi-direct vision odometer based on binocular vision | |
| CN105335934B (en) | Disparity map computational methods and device | |
| CN109697725B (en) | Background filtering method and device and computer readable storage medium | |
| US20130223725A1 (en) | Apparatus and method for estimating disparity using visibility energy model | |
| CN107564053B (en) | Parallax value correction method and device applied to road scene | |
| Emlek et al. | Variable window size for stereo image matching based on edge information | |
| CN116109669A (en) | Target tracking method and system and electronic equipment | |
| CN107204013B (en) | Method and device for calculating pixel point parallax value applied to binocular stereo vision | |
| EP3389015A1 (en) | Roll angle calibration method and roll angle calibration device | |
| CN107330932B (en) | Method and device for repairing noise in parallax map | |
| CN111967484A (en) | Point cloud clustering method and device, computer equipment and storage medium | |
| CN108022429B (en) | Vehicle detection method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |