A method for identifying a drivable area on an unstructured road, an electronic device and a medium

Abstract

The present invention discloses a method, electronic device and medium for identifying a drivable area on an unstructured road. The method comprises the following steps: converting the color channel of a road image; extracting a binary image based on color features; extracting a binary image based on texture features; merging binary images based on color features and texture features; obtaining the maximum connected domain of the image, and identifying the road outer enclosing convex polygon corresponding to the maximum connected domain of the image as a drivable area. The present invention simplifies the complex and diverse unstructured road pavement environment, performs boundary detection in combination with road color and boundary texture features, uses the maximum convex polygon to realize road boundary detection, and improves the boundary detection effect when the road boundary is unclear.

Description

Technical Field

[0001] The present invention relates to the technical field of road recognition, and in particular to a method for identifying a drivable area of ​​an unstructured road, an electronic device and a medium. Background Art

[0002] Road detection can be divided into structured roads and unstructured roads according to road conditions. Structured roads have lane lines, so lane line detection can be used to divide the drivable area of ​​the road. Unstructured roads are usually of various types and have complex surrounding environments. Road boundary detection is highly disturbed. Currently, there is no general detection method for unstructured roads. There are three main methods for unstructured road boundary detection. One is to detect road boundaries by analyzing the color and boundary texture features of the road. This method requires little prior knowledge and is insensitive to the road shape, but has high requirements for the road environment. It is easy to misidentify road boundaries when shadows, water marks, etc. exist. The other is to use the road model method to build a road model based on the prior knowledge of the road to detect road boundaries. This method is not robust and is more suitable for unstructured roads that are similar to the road model. In addition, road boundary detection based on deep neural networks has developed rapidly in recent years. By using a large number of sample sets to train neural network models, unstructured road segmentation is achieved. This method relies heavily on sample sets, and the boundary detection effect is usually poor for unstructured road types that are not included in the training set. Summary of the invention

[0003] In view of the problems existing in the prior art, the first object of the present invention is to provide a method for identifying a drivable area of ​​an unstructured road, which can reduce the influence of road environment factors on road recognition and ensure the accuracy of road recognition.

[0004] A second object of the present invention is to provide an electronic device.

[0005] A third object of the present invention is to provide a computer-readable medium.

[0006] To achieve the above object, the present invention provides a method for identifying a drivable area on an unstructured road in a first aspect, comprising the following steps:

[0007] Converting the RGB image corresponding to the road image captured by the camera into an HSV image according to a first predetermined formula;

[0008] According to the comparison result between the pixel value of the HSV image and the predetermined threshold, the image is divided into regions and a first image processing result about the color feature is output;

[0009] Extracting edge features of the road image by performing convolution on the RGB image in the horizontal and vertical directions;

[0010] Performing edge segmentation on the road image according to a comparison result between a pixel value of the image after edge extraction and a predetermined threshold value and outputting a second image processing result on texture features;

[0011] The binary image of the first image processing result and the binary image of the second image processing result are combined and calculated according to a predetermined formula to obtain a third image processing result of the road image;

[0012] The maximum connected domain of the image is obtained according to the binary image corresponding to the third image processing result, and the convex polygon surrounding the road corresponding to the maximum connected domain of the image is identified as a drivable area.

[0013] Further, the first predetermined formula is:

[0014]

[0015]

[0016]

[0017] Among them, R is the pixel value of the red channel, G is the pixel value of the green channel, B is the pixel value of the blue channel, H is the pixel value of the hue channel, S is the pixel value of the saturation channel, and V is the pixel value of the brightness channel.

[0018] Further, dividing the image into regions according to the comparison result between the pixel value of the HSV image and a predetermined threshold and outputting a first image processing result about the color feature includes:

[0019] When the pixel value of the HSV image is greater than a predetermined threshold, the image area corresponding to the pixel value is divided into a driving area;

[0020] When the pixel value of the HSV image is less than a predetermined threshold, the area corresponding to the pixel value is divided into a background area;

[0021] The driving area and background area of ​​the H, S, and V channels in the HSV image are binarized, and the first image processing result is output; wherein the first image processing result includes a binarized image corresponding to the H channel, a binarized image corresponding to the S channel, and a binarized image corresponding to the V channel.

[0022] Further, extracting edge features of the road image by performing convolution on the RGB image in the horizontal direction and the vertical direction includes:

[0023] Perform image processing on the RGB image to obtain a grayscale image;

[0024] After performing horizontal and vertical convolution on the grayscale image, the pixel values ​​of the image after horizontal convolution and the pixel values ​​of the image after vertical convolution are obtained.

[0025] Further, performing edge segmentation on the road image and outputting a second image processing result about texture features according to a comparison result between a pixel value of the image after edge extraction and a predetermined threshold value comprises:

[0026] When the pixel value of the image after edge extraction is greater than a predetermined threshold, the area corresponding to the pixel value is divided into an edge area;

[0027] When the pixel value of the image after edge extraction is less than a predetermined threshold, the area corresponding to the pixel value is divided into a non-edge area;

[0028] Binarize the edge area and non-edge area of ​​the image after horizontal convolution and the image after vertical convolution in the edge extracted image, and output the second image processing result; wherein the second image processing result includes the binarized image corresponding to the image after horizontal convolution and the binarized image corresponding to the image after vertical convolution.

[0029] Further, the binary image of the first image processing result and the binary image of the second image processing result are combined and calculated according to a predetermined formula to obtain a third image processing result of the road image, including:

[0030] The binary image corresponding to the H channel, the binary image corresponding to the S channel, and the binary image corresponding to the V channel are ORed to obtain a binary image corresponding to the color feature;

[0031] Perform an OR operation on the binary image corresponding to the image after the horizontal convolution and the binary image corresponding to the image after the vertical convolution to obtain a binary image corresponding to the texture feature;

[0032] The third image processing result is obtained by performing an AND operation on the binary image corresponding to the color feature and the binary image corresponding to the texture feature.

[0033] Further, obtaining the maximum connected domain of the image according to the binary image corresponding to the third image processing result, and identifying the road outer enclosing convex polygon corresponding to the maximum connected domain of the image as a drivable area includes:

[0034] Marking all non-zero pixels in the binary image corresponding to the third image processing result;

[0035] The convex polygon obtained by connecting the outermost points of the connected area with the largest number of non-zero pixels is identified as the drivable area.

[0036] Furthermore, it also includes:

[0037] The KCF algorithm is used to eliminate abnormal results in the road disappearance areas of the identified drivable areas.

[0038] A second aspect of the present invention provides an electronic device, comprising:

[0039] one or more processors; and

[0040] A storage device is used to store one or more programs. When the one or more programs are executed by the one or more processors, the one or more processors implement the drivable area identification method described in the first aspect above.

[0041] A third aspect of the present invention provides a computer-readable medium having a computer program stored thereon, wherein when the program is executed by a processor, the method for identifying a drivable area described in the first aspect is implemented.

[0042] The present invention simplifies the complex and diverse unstructured road pavement environment, combines road color and boundary texture features for boundary detection, uses the maximum convex polygon to achieve road boundary detection, and improves the boundary detection effect when the road boundary is unclear.

[0043] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0045] Figure 1 A schematic diagram of a flow chart of a method for identifying a drivable area on an unstructured road according to an embodiment of the present invention;

[0046] Figure 2-Figure 6 A schematic diagram of a flow chart of a method for identifying a drivable area on an unstructured road according to another embodiment of the present invention;

[0047] Figure 7 A schematic diagram of an image after the image is divided into regions according to the present invention;

[0048] Figure 8 This is a schematic diagram of an image of edge features of a road image extracted by convolution in the present invention;

[0049] Fig. 9 A schematic diagram of an image selected for the largest convex polygon of the present invention;

[0050] Fig.10 It is a schematic diagram of an image for tracking the road disappearing area according to the present invention. DETAILED DESCRIPTION

[0051] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided so that the present invention will be more comprehensive and complete and fully convey the concept of the example embodiments to those skilled in the art.

[0052] In addition, the described features, structures or characteristics may be combined in one or more embodiments in any suitable manner. In the following description, many specific details are provided to provide a full understanding of the embodiments of the present invention. However, those skilled in the art will appreciate that the technical solution of the present invention can be practiced without one or more of the specific details, or other methods, components, devices, steps, etc. may be adopted. In other cases, known methods, devices, implementations or operations are not shown or described in detail to avoid blurring various aspects of the present invention.

[0053] The block diagrams shown in the accompanying drawings are merely functional entities and do not necessarily correspond to physically independent entities. That is, these functional entities may be implemented in software form, or in one or more hardware modules or integrated circuits, or in different networks and / or processor devices and / or microcontroller devices.

[0054] The flowcharts shown in the accompanying drawings are only exemplary and do not necessarily include all the contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps can be decomposed, and some operations / steps can be combined or partially combined, so the actual execution order may change according to actual conditions.

[0055] like Figure 1 As shown, the first aspect of the present invention provides a method for identifying a drivable area on an unstructured road, comprising the following steps:

[0056] Step S110: converting the RGB image corresponding to the road image captured by the camera into an HSV image according to a first predetermined formula.

[0057] Step S120: dividing the image into regions according to the comparison result between the pixel value of the HSV image and a predetermined threshold value and outputting a first image processing result regarding color features.

[0058] Step S130: extracting edge features of the road image by performing convolution on the RGB image in the horizontal and vertical directions.

[0059] Step S140: performing edge segmentation on the road image according to the comparison result between the pixel value of the image after edge extraction and the predetermined threshold value, and outputting a second image processing result about texture features.

[0060] Step S150: merging the binarized image of the first image processing result and the binarized image of the second image processing result according to a predetermined formula to obtain a third image processing result of the road image.

[0061] Step S160: obtaining the maximum connected domain of the image according to the binary image corresponding to the third image processing result, and identifying the road outer enclosing convex polygon corresponding to the maximum connected domain of the image as a drivable area.

[0062] The present invention simplifies the complex and diverse unstructured road pavement environment, combines road color and boundary texture features for boundary detection, uses the maximum convex polygon to achieve road boundary detection, and improves the boundary detection effect when the road boundary is unclear.

[0063] In one embodiment of the present invention, the first predetermined formula for image conversion used in step S110 is:

[0064]

[0065]

[0066]

[0067] Among them, R is the pixel value of the red channel, G is the pixel value of the green channel, B is the pixel value of the blue channel, H is the pixel value of the hue channel, S is the pixel value of the saturation channel, and V is the pixel value of the brightness channel. Among them, the value range of R, G, and B is 0 to 255. H is the hue, and the saturation S indicates the degree to which the color is close to the spectral color. The larger the value, the more saturated the color. The brightness V indicates the brightness of the color. For the light source color, the brightness value is related to the brightness of the light source. For the object color, the brightness value is related to the transmittance or reflectance of the object. For unstructured roads, the light intensity and shadows have a greater impact on the recognition of the drivable area of ​​the road, while the H channel is less affected by shadows and brightness. Combined with the road information obtained by the S and V channels, using the HSV channel for drivable area recognition can obtain more stable and reliable results compared with the original RGB channel image. By converting the RGB channel into the HSV channel value, the images of the three channels of H, S, and V are obtained to realize the input image preprocessing.

[0068] like Figure 2 As shown, step S120 also includes steps S210 to S230:

[0069] Step S210: when the pixel value of the HSV image is greater than a predetermined threshold, the image area corresponding to the pixel value is divided into a driving area.

[0070] Step S220: when the pixel value of the HSV image is less than a predetermined threshold, the area corresponding to the pixel value is divided into a background area;

[0071] Step S230: Binarize the driving area and background area of ​​the H, S, and V channels in the HSV image, and output the first image processing result; wherein the first image processing result includes a binary image corresponding to the H channel, a binary image corresponding to the S channel, and a binary image corresponding to the V channel.

[0072] By setting the area with pixel values ​​greater than the preset threshold to 1 (i.e., the drivable area) and the area with pixel values ​​less than the threshold to 0 (i.e., the background area), the drivable area is divided and the binary image A corresponding to the H channel, the binary image B corresponding to the S channel and the binary image C corresponding to the V channel are obtained.

[0073] Since the color space threshold of unstructured roads changes continuously under different road sections and lighting conditions, the preset threshold of the present invention can be adaptive according to the color space change. For a channel in H, S, V of the image I (x, y), x and y are pixel coordinate values, and the coordinates correspond to pixel points in the range of 0 to 255. The segmentation threshold of the drivable area and the background is recorded as T, and the proportion of the number of pixels belonging to the foreground to the entire image is recorded as w. 0 , the corresponding pixel average value is u 0 , the ratio of background pixels to the entire image is w 1 , the corresponding pixel average value is u 1 , the average value of all pixels in the image is u, the variance between classes is g, the image size is M×N, and the number of pixels in the drivable area is N 0 , the number of background pixels is N 1 , then:

[0074] w 0 =N 0 / (M×N) (1)

[0075] w 1 =N 1 / (M×N) (2)

[0076] N 0 +N 1 =M×N (3)

[0077] w 0 +w 1 =1 (4)

[0078] μ=w 0 ×μ 0 +w 1 ×μ 1 (5)

[0079] g=w 0 (μ 0 -μ) 2 +w 1 (μ 1 -μ) 2 (6)

[0080] Substituting formula (5) into formula (6), we get the equivalent formula:

[0081] g=w 0 w 1 (μ 0 -μ 1 ) 2

[0082] The threshold T is traversed from 0 to 255. When the pixels of the drivable area and the background are misclassified, the inter-class variance will decrease. The larger the inter-class variance, the greater the difference between the two parts, and the smaller the probability of regional misclassification. The threshold T corresponding to the maximum inter-class variance g is the predetermined threshold. Figure 7 .

[0083] like Figure 3 As shown, step S130 also includes step S310 and step S320:

[0084] Step S310: performing image processing on the RGB image to obtain a grayscale image.

[0085] For the original RGB image, the grayscale image A is obtained by A=0.299×R+0.587×G+0.114×B.

[0086] Step S320: After performing horizontal and vertical convolution on the grayscale image, the pixel values ​​of the image after horizontal convolution and the pixel values ​​of the image after vertical convolution are obtained. Two 3×3 convolution factors are used to convolve the grayscale image A, that is, the image is convolved horizontally and vertically. After traversing the entire image, the pixel value of the image after horizontal convolution is G x After traversing the entire image and undergoing vertical convolution, the pixel value of the image is G y .

[0087]

[0088] Convolution is used to extract edge features of road images. Figure 8 .

[0089] like Figure 4As shown, step S140 also includes steps S410 to S430:

[0090] Step S410: when the pixel value of the image after edge extraction is greater than a predetermined threshold, the area corresponding to the pixel value is divided into an edge area.

[0091] Step S420: when the pixel value of the image after edge extraction is less than a predetermined threshold, the area corresponding to the pixel value is divided into a non-edge area.

[0092] Step S430: Binarize the edge area and non-edge area of ​​the image after horizontal convolution and the image after vertical convolution in the edge extracted image, and output the second image processing result; wherein the second image processing result includes the binarized image corresponding to the image after horizontal convolution and the binarized image corresponding to the image after vertical convolution.

[0093] After the image is subjected to horizontal and vertical edge extraction, it is binarized using the same adaptive threshold, with the edge area set to 0 and the remaining area set to 1, to obtain the binary image D and E corresponding to the image after horizontal convolution and the image after vertical convolution.

[0094] like Figure 5 As shown, step S150 also includes steps S510 to S530:

[0095] Step S510: performing an OR operation on the binary image corresponding to the H channel, the binary image corresponding to the S channel, and the binary image corresponding to the V channel to obtain a binary image corresponding to the color feature.

[0096] Step S520: performing an OR operation on the binary image corresponding to the image after the horizontal convolution and the binary image corresponding to the image after the vertical convolution to obtain a binary image corresponding to the texture feature.

[0097] Step S530: performing an AND operation on the binary image corresponding to the color feature and the binary image corresponding to the texture feature to obtain the third image processing result.

[0098] A total of 5 binarization results are obtained by using color features and texture features. The 3 binarization results obtained by using color features are A, B, and C, and the 2 binarization results obtained by using texture features are D and E.

[0099] A single binary image Z is obtained by using the following OR operation and AND operation, where “|” is an OR operation and “&” is an AND operation.

[0100] X=A|B|C

[0101] Y=D|E

[0102] Z=X&Y

[0103] The obtained binary image Z roughly identified the drivable area through color features, and accurately identified the boundary of the drivable area through texture features. However, in the merged result, the cracks in the middle of the road and the weeds at the road boundary affected the result. In the final result, part of the drivable area was mistakenly set to 0, and some pixels outside the boundary of the drivable area were also mistakenly set to 1. The impact of weeds on the road boundary can be improved through further corrosion and expansion processing. For the interference of debris in the middle of the road, the maximum convex polygon method is used to further correct the result.

[0104] like Figure 6 As shown, step S160 also includes step S610 and step S620:

[0105] Step S610: Mark all non-zero pixels in the binarized image corresponding to the third image processing result.

[0106] Step S620: A convex polygon obtained by connecting the outermost points of the connected area with the largest number of non-zero pixels is identified as a drivable area.

[0107] In order to reduce the impact of cracks, shadows, and debris in the middle of the road on the identification of road boundaries, combined with the perspective deformation of camera imaging with large near and small far, the drivable area of ​​the road is usually approximately trapezoidal. The present invention uses color features and texture features to extract the input image to obtain the binary drivable area and background, and uses the binary image to obtain the maximum connected domain and the corresponding outer bounding convex multi-deformation, which is the drivable area.

[0108] Specifically, the following steps are included:

[0109] ① Extract the input image using color features and texture features to obtain the binary drivable area and background;

[0110] ②Add an attribute to the binary image to indicate whether it has been visited;

[0111] ③ Find the first non-zero pixel, push it into the stack and set its visited attribute to true;

[0112] ④ Taking whether the stack size is 0 as the end condition, find the eight neighboring non-zero pixels adjacent to the top element of the stack, push them into the stack, and delete the top element of the stack after completion;

[0113] ⑤When the stack is empty, it means that a connected area has been traversed, and it is necessary to continue to find the next non-empty and unvisited pixel as the starting point, and repeat the above steps ① to ④ until all non-zero pixels have been visited;

[0114] ⑥ After all connected regions are solved, mark the connected region with the largest number of pixels and use OpenCV to obtain the outermost points of the largest connected domain. Connect the points to form a convex polygon, that is, the largest convex polygon formed by the largest connected domain. See the image of the largest convex polygon for selection. Fig. 9 .

[0115] In one embodiment of the present invention, the method for identifying a drivable area on an unstructured road further includes:

[0116] The KCF algorithm is used to eliminate abnormal results in the road disappearance areas of the identified drivable areas.

[0117] In the same continuous unstructured road, except for a few sections, the road width usually does not change drastically. Therefore, based on this prior knowledge, outliers can be initially eliminated according to the convex polygon area formed by the road area and the road boundary slope. When the convex polygon or boundary slope changes suddenly due to shadows or other factors, it is regarded as an outlier and no result is output. In addition, since the road has the advantage of obvious small area feature differences in the area where the field of vision disappears, the method of manually selecting the rectangular area where the road disappears is adopted. For tracking the road disappearing area, see Fig.10 The KCF algorithm is used to collect positive and negative samples using the circulant matrix of the target area, and the target detector is trained using ridge regression to achieve tracking of the road disappearing area. During the tracking process, the coordinate change value of the center point of the road disappearing rectangular area is output. Since this value can be output stably and accurately, it can be used as a reference value for the direction of change of the road boundary slope. When the direction of change of the center point coordinate is inconsistent with the direction of change of the road boundary slope, detection anomalies usually occur, so KCF tracking can achieve partial anomaly detection elimination.

[0118] A second aspect of the present invention provides an electronic device, comprising:

[0119] one or more processors; and

[0120] A storage device is used to store one or more programs. When the one or more programs are executed by the one or more processors, the one or more processors implement the drivable area identification method described in the first aspect above.

[0121] A third aspect of the present invention provides a computer-readable medium having a computer program stored thereon, wherein when the program is executed by a processor, the method for identifying a drivable area described in the first aspect is implemented.

[0122] According to an embodiment of the present invention, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present invention includes a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through a communication part, and / or installed from a removable medium. When the computer program is executed by a central processing unit (CPU), the above-mentioned functions defined in the system of the present application are executed.

[0123] It should be noted that the computer-readable medium shown in the present invention may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present invention, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, device or device. In the present invention, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable program code. This propagated data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than computer-readable storage media, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer-readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wire, optical cable, RF, etc., or any suitable combination of the above.

[0124] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present invention. In this regard, each box in the flow chart or block diagram can represent a module, a program segment, or a part of a code, and the above-mentioned module, program segment, or a part of a code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flow chart, and the combination of the boxes in the block diagram or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0125] The modules involved in the embodiments of the present invention may be implemented by software or hardware, and the modules described may also be arranged in a processor. The names of these modules do not, in some cases, limit the modules themselves.

[0126] It should be noted that, although several modules or units of the equipment for action execution are mentioned in the above detailed description, this division is not mandatory. In fact, according to an embodiment of the present invention, the features and functions of two or more modules or units described above can be embodied in one module or unit. On the contrary, the features and functions of one module or unit described above can be further divided into being embodied by multiple modules or units.

[0127] Through the description of the above implementation, it is easy for those skilled in the art to understand that the example implementation described here can be implemented by software, or by software combined with necessary hardware. Therefore, the technical solution according to the implementation of the present invention can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, including several instructions to enable a computing device (which can be a personal computer, a server, a touch terminal, or a network device, etc.) to execute the method according to the implementation of the present invention.

[0128] Those skilled in the art will readily appreciate other embodiments of the present invention after considering the specification and practicing the invention disclosed herein. This application is intended to cover any variations, uses or adaptations of the present invention that follow the general principles of the present invention and include common knowledge or customary techniques in the art that are not disclosed by the present invention. The specification and examples are to be considered exemplary only, and the true scope and spirit of the present invention are indicated by the following claims.

[0129] It should be understood that the present invention is not limited to the exact construction that has been described above and shown in the drawings and that various modifications and changes may be made without departing from the scope thereof. The scope of the present invention is limited only by the appended claims.

Claims

1. A method for identifying a drivable area on an unstructured road, characterized in that: The steps include: The RGB image corresponding to the road image captured by the camera is converted into an HSV image according to a first predetermined formula; the first predetermined formula is: Among them, R is the pixel value of the red channel, G is the pixel value of the green channel, B is the pixel value of the blue channel, H is the pixel value of the hue channel, S is the pixel value of the saturation channel, and V is the pixel value of the brightness channel; According to the comparison result between the pixel value of the HSV image and the predetermined threshold, the image is divided into regions and a first image processing result about the color feature is output; Extracting edge features of the road image by performing convolution on the RGB image in the horizontal and vertical directions; Performing edge segmentation on the road image according to a comparison result between a pixel value of the image after edge extraction and a predetermined threshold value and outputting a second image processing result on texture features; The binary image of the first image processing result and the binary image of the second image processing result are combined and calculated according to a predetermined formula to obtain a third image processing result of the road image; Obtaining the maximum connected domain of the image according to the binary image corresponding to the third image processing result, and identifying the road outer enclosing convex polygon corresponding to the maximum connected domain of the image as a drivable area; Extracting edge features of the road image by performing convolution on the horizontal and vertical directions of the RGB image includes: Perform image processing on the RGB image to obtain a grayscale image; After performing horizontal and vertical convolution on the grayscale image, the pixel values ​​of the image after horizontal convolution and the pixel values ​​of the image after vertical convolution are obtained; The method of dividing the road image by edge segmentation according to the comparison result between the pixel value of the image after edge extraction and the predetermined threshold value and outputting the second image processing result about the texture feature comprises: When the pixel value of the image after edge extraction is greater than a predetermined threshold, the area corresponding to the pixel value is divided into an edge area; When the pixel value of the image after edge extraction is less than a predetermined threshold, the area corresponding to the pixel value is divided into a non-edge area; Binarize the edge area and non-edge area of ​​the image after horizontal convolution and the image after vertical convolution in the edge extracted image, and output the second image processing result; wherein the second image processing result includes the binarized image corresponding to the image after horizontal convolution and the binarized image corresponding to the image after vertical convolution; Combining and calculating the binarized image of the first image processing result and the binarized image of the second image processing result according to a predetermined formula to obtain a third image processing result of the road image includes: The binary image corresponding to the H channel, the binary image corresponding to the S channel, and the binary image corresponding to the V channel are ORed to obtain a binary image corresponding to the color feature; Perform an OR operation on the binary image corresponding to the image after the horizontal convolution and the binary image corresponding to the image after the vertical convolution to obtain a binary image corresponding to the texture feature; The third image processing result is obtained by performing an AND operation on the binary image corresponding to the color feature and the binary image corresponding to the texture feature.

2. The method for identifying a drivable area on an unstructured road according to claim 1, characterized in that: The method of dividing the image into regions according to the comparison result between the pixel value of the HSV image and the predetermined threshold and outputting the first image processing result about the color feature comprises: When the pixel value of the HSV image is greater than a predetermined threshold, the image area corresponding to the pixel value is divided into a driving area; When the pixel value of the HSV image is less than a predetermined threshold, the area corresponding to the pixel value is divided into a background area; The driving area and background area of ​​the H, S, and V channels in the HSV image are binarized, and the first image processing result is output; wherein the first image processing result includes a binarized image corresponding to the H channel, a binarized image corresponding to the S channel, and a binarized image corresponding to the V channel.

3. The method for identifying a drivable area on an unstructured road according to claim 1, characterized in that: Obtaining the maximum connected domain of the image according to the binary image corresponding to the third image processing result, and identifying the road outer enclosing convex polygon corresponding to the maximum connected domain of the image as a drivable area includes: Marking all non-zero pixels in the binary image corresponding to the third image processing result; The convex polygon obtained by connecting the outermost points of the connected area with the largest number of non-zero pixels is identified as the drivable area.

4. The method for identifying a drivable area on an unstructured road according to claim 1, wherein: Also includes: The KCF algorithm is used to eliminate abnormal results in the road disappearance area of ​​the identified drivable area.

5. An electronic device, characterized in that: include: one or more processors; as well as A storage device for storing one or more programs, which, when executed by the one or more processors, enables the one or more processors to implement the drivable area identification method as described in any one of claims 1 to 4.

6. A computer readable medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the drivable area identification method according to any one of claims 1 to 4 is implemented.

Images