CN113420682B

CN113420682B - Target detection method and device in vehicle-road cooperation and road side equipment

Info

Publication number: CN113420682B
Application number: CN202110721853.4A
Authority: CN
Inventors: 夏春龙
Original assignee: Apollo Zhilian Beijing Technology Co Ltd
Current assignee: Apollo Zhilian Beijing Technology Co Ltd; Apollo Zhixing Technology Guangzhou Co Ltd
Priority date: 2021-06-28
Filing date: 2021-06-28
Publication date: 2023-08-15
Anticipated expiration: 2041-06-28
Also published as: JP2023536025A; WO2023273041A1; KR20220091607A; CN113420682A; JP7436670B2

Abstract

The invention discloses a target detection method, device and drive test equipment in vehicle-road cooperation, relates to the field of intelligent transportation, and particularly relates to the technical field of image detection. The specific implementation scheme is as follows: performing target detection on an image to obtain a candidate target region in the image, the confidence level of the candidate target region and the blocked level of the candidate target region; updating the confidence coefficient of the candidate target region based on the intersection ratio between the candidate target regions and the blocked degree of the candidate target region; and detecting the target in the image from the candidate target area according to the updated confidence. When the scheme provided by the embodiment of the disclosure is applied to target detection, the accuracy of target detection is improved.

Description

Target detection method and device in vehicle-road cooperation and road side equipment

Technical Field

The disclosure relates to the technical field of intelligent transportation, in particular to the technical field of image detection.

Background

In application scenes such as road monitoring, vehicle path planning and the like of the vehicle-road cooperation V2X, after the image acquired by the image acquisition equipment is obtained, targets such as people, animals, vehicles and the like in the image are required to be detected so as to locate the targets in the image, and then the processing operation for the targets is triggered, or the vehicle path planning and the like are carried out by combining the targets. Therefore, there is a need for a method of detecting objects in vehicle-road collaboration to detect objects in images.

Disclosure of Invention

The disclosure provides a method and a device for detecting targets in vehicle-road cooperation and drive test equipment.

According to an aspect of the present disclosure, there is provided a method for detecting a target in vehicle-road cooperation, the method including:

performing target detection on an image to obtain a candidate target region in the image, the confidence level of the candidate target region and the blocked level of the candidate target region;

updating the confidence coefficient of the candidate target region based on the intersection ratio between the candidate target regions and the blocked degree of the candidate target region;

and detecting the target in the image from the candidate target area according to the updated confidence.

According to an aspect of the present disclosure, there is provided a device for detecting an object in vehicle-road cooperation, the device including:

the information acquisition module is used for carrying out target detection on the image to obtain a candidate target area in the image, the confidence level of the candidate target area and the shielding level of the candidate target area;

the confidence updating module is used for updating the confidence of the candidate target areas based on the intersection ratio between the candidate target areas and the blocked degree of the candidate target areas;

and the target detection module is used for detecting the target in the image from the candidate target area according to the updated confidence level.

According to another aspect of the present disclosure, there is provided an electronic device including:

at least one processor; and

a memory communicatively coupled to the at least one processor; wherein,,

the memory stores instructions executable by the at least one processor to enable the at least one processor to implement a method of target detection in vehicle-road coordination.

According to another aspect of the present disclosure, there is provided a non-transitory computer-readable storage medium storing computer instructions for causing the computer to perform a method of target detection in vehicle-road collaboration.

According to another aspect of the present disclosure, a computer program product is provided, comprising a computer program which, when executed by a processor, implements a method of target detection in vehicle-road coordination.

According to another aspect of the present disclosure, there is provided a roadside apparatus including the above electronic apparatus.

According to another aspect of the present disclosure, a cloud control platform is provided, including the above electronic device.

From the above, when the scheme provided by the embodiment of the present disclosure is applied to target detection, the confidence level of the candidate target region is updated according to the intersection ratio between the candidate target regions and the blocked level of the candidate target region, and then the target in the image is detected from the candidate target region based on the updated confidence level. Because the intersection ratio between the candidate target areas can reflect the overlapping ratio between the candidate target areas, the blocked degree of the candidate target areas can reflect the blocked degree of the candidate target areas, the confidence level is updated according to the intersection ratio and the blocking degree, so that the confidence level of the target areas can refer to the overlapping condition between the target areas, the updated confidence level of the candidate target areas is more prone to the actual condition, the image is subjected to target detection according to the updated confidence level, and the accuracy of target detection can be improved.

It should be understood that the description in this section is not intended to identify key or critical features of the embodiments of the disclosure, nor is it intended to be used to limit the scope of the disclosure. Other features of the present disclosure will become apparent from the following specification.

Drawings

The drawings are for a better understanding of the present solution and are not to be construed as limiting the present disclosure. Wherein:

fig. 1 is a schematic flow chart of a method for detecting targets in vehicle-road cooperation according to an embodiment of the disclosure;

FIG. 2 is a schematic illustration of an image provided in accordance with an embodiment of the present disclosure;

FIG. 3a is a schematic diagram of a network model provided in accordance with an embodiment of the present disclosure;

FIG. 3b is a schematic diagram of another network model provided in accordance with an embodiment of the present disclosure;

fig. 4 is a schematic structural diagram of a target detection device in vehicle-road cooperation according to an embodiment of the present disclosure;

fig. 5 is a schematic structural view of an electronic device provided according to an embodiment of the present disclosure.

Detailed Description

Exemplary embodiments of the present disclosure are described below in conjunction with the accompanying drawings, which include various details of the embodiments of the present disclosure to facilitate understanding, and should be considered as merely exemplary. Accordingly, one of ordinary skill in the art will recognize that various changes and modifications of the embodiments described herein can be made without departing from the scope and spirit of the present disclosure. Also, descriptions of well-known functions and constructions are omitted in the following description for clarity and conciseness.

The embodiment of the disclosure provides a method and a device for detecting targets in vehicle-road cooperation and road test equipment.

In one embodiment of the present disclosure, a method for detecting a target in vehicle-road coordination is provided, including:

performing target detection on the image to obtain a candidate target region in the image, the confidence level of the candidate target region and the blocked level of the candidate target region;

Because the intersection ratio between the candidate target areas can reflect the overlapping ratio between the candidate target areas, the blocked degree of the candidate target areas can reflect the blocked degree of the candidate target areas, the confidence level is updated according to the intersection ratio and the blocking degree, so that the confidence level of the target areas can refer to the overlapping condition between the target areas, the updated confidence level of the candidate target areas is more prone to the actual condition, the image is subjected to target detection according to the updated confidence level, and the accuracy of target detection can be improved.

The execution bodies of the embodiments of the present disclosure are described below.

The execution body of the embodiment of the present disclosure may be an electronic device integrated with an object detection function, where the electronic device may be: desktop computers, notebook computers, servers, image acquisition devices, and the like. Wherein, the image acquisition device may include: video cameras, automobile recorders, etc.

The scheme provided by the embodiment of the disclosure can be applied to target detection of the images acquired in application scenes such as road monitoring of the vehicle-road cooperation V2X, vehicle path planning and the like.

In addition, the scheme provided by the embodiment of the disclosure can also be used for detecting the targets of the images acquired in other scenes. For example, the other scenes may be highly dense scenes of persons such as subway stations, markets, singing concerts, etc., and the images of the scenes are collected, so that people tend to be dense in the collected images, and the situation that the faces of some people are blocked by the faces of other people easily occurs. The scene can also be a scene with dense personnel at the entrance of a museum, a bank hall and the like, and the scene is subjected to image acquisition, so that the situation that the face of the personnel is blocked by other personnel or buildings and the like possibly occurs in the acquired image.

The foregoing is merely an example of an application scenario of the embodiments of the present disclosure, and is not intended to limit the present disclosure.

The object may be a human face, an animal, a vehicle, etc.

The method for detecting the targets in the vehicle-road cooperation provided by the embodiment of the disclosure is specifically described below.

Referring to fig. 1, fig. 1 is a flow chart of a method for detecting a target in vehicle-road cooperation according to an embodiment of the disclosure, where the method includes the following steps S101 to S103.

Step S101: and carrying out target detection on the image to obtain a candidate target region in the image, the confidence level of the candidate target region and the shielding level of the candidate target region.

The image may be obtained by image acquisition for a specific scene. The scene may include a vehicle driving scene, a parking lot scene, in which case the target is a vehicle; the scene may also include public space scenes such as subway stations, high-speed rail stations, etc., in which case the target is a person.

For different types of targets, different detection algorithms can be adopted to realize target detection, such as a face detection algorithm, a license plate detection algorithm and the like.

The candidate target area refers to: the area where the target is considered likely to exist is detected by the target. Taking fig. 2 as an example, the region surrounded by the rectangular frames in fig. 2 is a candidate target region obtained by performing animal detection on the image.

Confidence of candidate target region reflects: the likelihood size of the target in the candidate target area. The confidence may be expressed in decimal, percent, etc. The greater the confidence value, the higher the likelihood that a target exists in the candidate target region.

For example: in the case of the target being a person, when the confidence of the candidate target area a is greater than the confidence of the candidate target area B, this indicates that the likelihood of the person being present in the candidate target area a is higher than the likelihood of the person being present in the candidate target area B.

The occlusion degree of the candidate target region reflects: the extent to which the candidate target region is occluded. The above-mentioned blocked degree can be expressed in decimal, percentage, etc., can also be expressed in the number of blocked grades, blocked grades can include serious blocking, moderate blocking, slight blocking, etc., blocked grades number includes 1, 2, 3, etc.

The specific target detection process may be found in the following examples, which are not described in detail herein.

Step S102: and updating the confidence of the candidate target region based on the intersection ratio between the candidate target regions and the occlusion degree of the candidate target region.

The intersection ratio between candidate target areas is used to describe the degree of overlap between two candidate target areas.

Specifically, the overlapping area between two candidate target areas can be calculated to obtain a first area, the sum of the areas of the two candidate target areas is calculated to obtain a second area, then the difference between the second area and the first area is calculated to obtain a third area, and the ratio between the first area and the third area is determined as the intersection ratio between the candidate target areas.

For example: the area of the candidate target area a is 48, the area of the candidate target area B is 32, wherein the overlapping area of the candidate target area a and the candidate target area B is 16, that is, the first area is 16, the total area of the candidate target area a and the candidate target area B is (46+32) =80, that is, the second area is 80, the difference (80-16) =64 between the second area and the first area is calculated, that is, the third area is 64, the ratio of the first area and the third area is calculated to obtain 16/64=0.25, and 0.25 is the cross-over ratio between the candidate target areas.

In one implementation, a reference region may be selected from among the candidate target regions, the intersection ratio of each other candidate target region to the reference region may be calculated, and the calculated intersection ratio may be determined as the intersection ratio for updating the confidence of the candidate target region. For example: the region with the highest confidence level may be selected from among the candidate target regions as the reference region.

In another implementation, for each candidate target region, a cross-over ratio may be selected from the cross-over ratios between the candidate target region and other candidate target regions, and the selected cross-over ratio may be determined as the cross-over ratio used to update the confidence of the candidate target region. For example: the maximum cross-over ratio, the average cross-over ratio, the median cross-over ratio, or the minimum cross-over ratio may be selected from the plurality of cross-over ratios.

When updating the confidence coefficient of the candidate target area, the adjustment coefficient can be calculated according to the intersection ratio between the candidate target areas and the blocked degree of the candidate target area and the preset first weight and the second weight, and the confidence coefficient of the candidate target area is updated according to the calculated adjustment coefficient.

Specifically, the product of the intersection ratio between the candidate target areas and the first weight can be calculated, the product of the blocked degree of the candidate target areas and the second weight can be calculated, and the sum of the calculated products is used as an adjustment coefficient.

For example: the intersection ratio between candidate target areas is 80%, the shielding degree of the candidate target areas is 50%, the preset first weight is 0.8, the preset second weight is 0.2, and the product of the intersection ratio between the candidate target areas and the first weight is calculated as follows: 0.8 x 80% = 64%, calculating the product between the occlusion degree of the candidate target region and the second weight as: 0.2 x 50% = 10%, the sum of the two calculated products is: 64% +10% = 74%, resulting in an adjustment coefficient of 74%.

After the adjustment coefficient is calculated, a product between the adjustment coefficient and the confidence coefficient of the candidate target region may be calculated as the updated confidence coefficient of the candidate target region.

Step S103: and detecting the target in the image from the candidate target area according to the updated confidence.

In one embodiment of the present disclosure, candidate target regions with updated confidence greater than a preset confidence threshold may be selected, and targets in the selected candidate target regions may be determined as targets in the image.

The preset confidence threshold may be set empirically by a worker, for example: the preset confidence threshold may be 90%,95%, etc. when the confidence is expressed as a percentage.

Describing the above-described target determination process by way of example, it is assumed that the confidence levels of the updated candidate target regions are respectively: 80%, 70%, 90%,95%, the preset confidence threshold is 85%, the updated confidence of greater than 85% is 90%,95%, wherein the updated confidence of region 1 is 90%, and the updated confidence of region 2 is 95%, so the objects in region 1 and the objects in region 2 are objects in the image.

Thus, for candidate target regions with a confidence greater than the preset confidence threshold, the likelihood of including targets in those candidate target regions is higher than the likelihood of including targets in other candidate target regions. Therefore, the targets in the candidate target areas with the confidence degrees larger than the preset confidence degree threshold value are determined to be targets in the image, and the obtained targets are high in accuracy.

In one embodiment of the present disclosure, a preset number of candidate target areas with the highest confidence after updating may also be selected, and the target in the selected candidate target areas is determined as the target in the image.

The above-mentioned preset number may be set empirically by a worker, for example: the preset number may be 1, 3, 5, etc.

The above-described targeting procedure is described as an example. The confidence of the target region is assumed to be respectively: 80%, 70%, 90%, 95%, 3 preset numbers, wherein the 3 largest confidence levels after updating are 95%, 90%, 80%, respectively; and determining the targets in the candidate target areas with the updated confidence degrees of 95%, 90% and 80% as targets in the image.

Thus, for a predetermined number of candidate target regions with a maximum confidence, the likelihood of including a target in those candidate target regions is higher than the likelihood of including a target in other candidate regions. Therefore, the targets in the preset number of candidate target areas with the maximum confidence are determined as the targets in the image, and the accuracy of the obtained targets is high.

From the above, when the scheme provided by the embodiment of the present disclosure is applied to target detection, the confidence level of the candidate target region is updated according to the intersection ratio between the candidate target regions and the blocked level of the candidate target region, and then the target in the image is detected from the candidate target region based on the updated confidence level. Because the intersection ratio between the candidate target areas can reflect the overlapping ratio between the candidate target areas, the blocked degree of the candidate target areas can reflect the blocked degree of the candidate target areas, the confidence level is updated according to the intersection ratio and the blocked degree, so that the confidence level of the target areas can refer to the overlapping condition between the target areas, the updated confidence level of the candidate target areas is more prone to the actual condition, the image is subjected to target detection according to the updated confidence level, and the accuracy of target detection can be improved.

In addition, the situation that the target is blocked is particularly serious in dense scenes, such as dense people flow and dense traffic flow scenes. For the images of the dense scenes, the occlusion degree of each candidate target area is higher, so that the targets in the candidate target areas are incomplete, and the confidence coefficient error of the obtained candidate target areas is larger. The confidence coefficient of the candidate target area is updated through the blocked degree of the candidate target area, so that the error influence on the confidence coefficient when each candidate target area is blocked can be effectively eliminated, the accuracy of the updated confidence coefficient is high, and an accurate target is detected. Therefore, the scheme provided by the embodiment of the disclosure can be better suitable for the condition that shielding exists in a dense scene, and the accuracy of target detection is improved.

In order to accurately update the confidence coefficient of the candidate target region, in one embodiment of the disclosure, a first region with the highest confidence coefficient may be selected from the region set in a circulating manner, and the confidence coefficient of other regions is updated according to the intersection ratio between other regions in the region set and the first region and the blocked degree of other regions, so that a confidence coefficient updating operation is completed, and the above operation is executed in a circulating manner until the region set includes one region. Each confidence update operation may be referred to as a loop.

The region set includes: unselected areas in the candidate target area. Specifically, at the beginning of the first cycle, the region set includes each candidate target region obtained in step S101; after selecting the first region from the region set in each cycle, the region set no longer includes the selected first region.

At the beginning of the first cycle, the first region is: the regions with the highest confidence in the candidate target regions obtained in the step S101; at each subsequent cycle, the first region is: the region with highest confidence in each updated region obtained after the last cycle

The other regions mentioned above refer to: the regions are concentrated in regions other than the first region. The region set includes: region 1, region 2, and region 3, wherein region 1 is the first region, regions other than the first region are region 2, region 3, and then region 2, region 3 are other regions.

And when each cycle is performed, traversing each region in the region set, sequencing the confidence of each region from high to low, and determining the region with the highest confidence as the first region. In addition, the first region may be stored in the prediction set, and as the number of loops increases, the number of first regions stored in the prediction set increases.

The above-described cyclic process is described below with reference to specific examples.

Let b1, b2, b3, … … bn be the respective candidate target areas obtained in step S101.

At the beginning of the first cycle, region set b= { B1, B2, B3, … …, bn }. Among these, the region with the highest confidence among the candidate target regions is the region b1, and therefore the region b1 is the first region. The regions other than the region B1 in the region set B are B2, B3, … …, bn, so { B2, B3, … …, bn } are other regions.

The confidence of the other regions { b2, b3, … …, bn } is updated based on the intersection ratio between the other regions { b2, b3, … …, bn } and the region b1 and the occlusion degree of the other regions { b2, b3, … …, bn }. And the first region may be added to the prediction set D, the added prediction set d= { b1};

at the beginning of the second cycle, since region B1 has been selected as the first region, region B1 is not included in region set B, region set b= { B2, B3, … …, bn }. Among these, the area with the highest confidence in the updated { b2, b3, … …, bn } is the area b2, and therefore the area b2 is the first area. The regions other than the region B2 in the region set B are B3, … …, bn, so { B3, … …, bn } are other regions.

The confidence of the other regions { b3, … …, bn } is updated based on the intersection ratio between the other regions { b3, … …, bn } and the region b2 and the occlusion degree of the other regions { b3, … …, bn }. And the first region may be added to the prediction set D, the added prediction set d= { b1, b2}.

At the beginning of the third cycle, since the regions B1, B2 have been selected as the first region, the region set b= { B3, … …, bn }. Among these, the area with the highest confidence in the updated { b3, … …, bn } is the area b3, and therefore the area b3 is the first area. The regions other than the region B3 in the region set B are B4, … …, bn, so { B4, … …, bn } are other regions.

The confidence of the other regions { b4, … …, bn } is updated based on the intersection ratio between the other regions { b4, … …, bn } and the region b3 and the occlusion degree of the other regions { b4, … …, bn }. And the first region may be added to the prediction set D, the added prediction set d= { b1, b2, b3}.

In a similar manner, until the number of the regions in the region set B is 1, directly adding the only one region in the region set B into the prediction set D, and obtaining the confidence of each updated region after the circulation is finished.

Thus, the confidence of the region in the region set is updated according to the intersection ratio of other regions in the region set and the first region and the occlusion degree of other regions in each cycle. The degree of shielding of other areas reflects the degree of shielding of other areas, the accuracy of the confidence of the detected area is low when the areas are shielded, and the accuracy of the confidence of the updated candidate target area can be high by introducing the degree of shielding of other areas; the intersection ratio between other areas and the first area reflects the coincidence ratio between the other areas and the first area, and the first area is the area with the highest confidence, and the confidence of the other areas can be effectively adjusted through the coincidence ratio between the first area and the area with the highest confidence. Therefore, the confidence of other areas can be effectively updated according to the intersection ratio and the occlusion degree in each cycle. And the accuracy of the updated confidence coefficient can be further improved through the cyclic iteration updating process.

In one embodiment of the present disclosure, the confidence of other regions may be updated as per steps A1-A4 below.

Step A1: and calculating the intersection ratio between other areas in the area set and the first area.

Specifically, firstly, calculating the overlapping area between other areas and the first area, and calculating the total area of the other areas and the first area; and then calculating the difference between the total area and the overlapping area to obtain a target area, and finally determining the ratio of the overlapping area to the target area as the intersection ratio between candidate target areas.

Step A2: and determining a first confidence coefficient adjusting value according to the cross-over ratio and a preset cross-over ratio threshold value.

The preset overlap ratio threshold may be set empirically by a worker, for example: the cross ratio threshold may be 90%, 95%, etc.

Specifically, in one embodiment, whether the intersection ratio is smaller than a preset intersection ratio threshold value may be determined, if yes, the first confidence adjustment value is determined to be a first preset value, and if not, the first confidence adjustment value is determined to be a second preset value.

The first preset value and the second preset value are set by staff according to experience.

In one embodiment of the present disclosure, it may also be determined whether the intersection ratio is less than a preset intersection ratio threshold; if yes, determining that the first confidence coefficient adjusting value is 1; if not, determining that the first confidence adjustment value is: 1 and the cross ratio.

For example: assuming that the preset intersection ratio threshold value is 90%, when the intersection ratio between other areas and the first area is 95%, determining that the intersection ratio 95% is greater than the preset intersection ratio threshold value of 90%, and determining a first confidence adjustment value: 1-95% = 5%; when the intersection ratio between the other areas and the first area is 50%, determining a first confidence coefficient adjusting value, wherein the intersection ratio 50% is smaller than a preset intersection ratio threshold value of 90%: 1.

in this way, when the intersection ratio is smaller than the preset intersection ratio threshold value, the overlapping ratio between the other areas and the first area is smaller, which means that the small part of the image content in the other areas is blocked, and the accuracy of the confidence coefficient of the other areas obtained by detection is high, and in this case, the confidence coefficient of the other areas can not be adjusted. Setting the first confidence adjustment value to 1 can achieve that the confidence of the region is not adjusted. When the intersection ratio is not smaller than the preset intersection ratio threshold value, the intersection ratio is larger, that is, the overlap ratio between other areas and the first area is larger, so that most of image contents in the other areas are blocked, the accuracy of the confidence coefficient of the other areas obtained through detection is low, and in this case, the confidence coefficient of the other areas needs to be adjusted. The first confidence coefficient adjusting value is set to be 1-the difference between the cross ratios, so that the adjusted confidence coefficient approaches to the actual situation.

Step A3: and determining a second confidence adjustment value according to the occluded degree of the other region.

In one embodiment, the product between the occlusion degree of the other region and the preset adjustment coefficient may be calculated as the second confidence adjustment value.

The above-mentioned preset adjustment coefficient may be set empirically by a worker, for example: the preset adjustment coefficient may be 1.2, 1.5, etc.

In one embodiment of the present disclosure, the second confidence adjustment value g (occ _pred) may also be determined according to the following expression:

g(occ_pred)＝α ^occ_pred

wherein occ _pred is the degree of occlusion of other areas, α is a preset constant, α >1.

Because α >1, the second confidence adjustment value g (occ _pred) increases as the occlusion of other regions increases.

When the occlusion degree of the region is high, the confidence degree of the region is low in accuracy, so that the confidence degree of the region needs to be greatly adjusted, and the adjusted confidence degree approaches to the actual situation. The second confidence coefficient adjustment value g (occ _pred) is increased along with the increase of the blocked degree of other areas, namely, the higher the blocked degree of other areas is, the larger the second confidence coefficient adjustment value is, so that the confidence coefficient of the other areas can be greatly adjusted, and the confidence coefficient of the adjusted other areas is close to the actual situation.

Step A4: and adjusting the confidence coefficient of other areas by adopting the first confidence coefficient adjusting value and the second confidence coefficient adjusting value.

In one embodiment of the present disclosure, the confidence of other regions may be adjusted according to the following expression:

S’＝S*T1*T2

wherein S' represents the confidence of the other region after adjustment, S represents the confidence of the other region before adjustment, T1 represents the first confidence adjustment value, and T2 represents the second confidence adjustment value.

In this way, the adjusted confidence coefficient is the product of the first confidence coefficient adjusting value, the second confidence coefficient adjusting value and the confidence coefficient of other areas, and the first confidence coefficient adjusting value and the second confidence coefficient adjusting value reflect the blocked condition of other areas from different angles, so that the adjusted confidence coefficient refers to the blocked condition of other areas, and the adjusted confidence coefficient is more approximate to the actual condition.

In another embodiment, the product of the first confidence coefficient adjustment value, the second confidence coefficient adjustment value and the confidence coefficient of other areas can be calculated and used as the reference confidence coefficient, the reference confidence coefficient is adjusted through a preset confidence coefficient error value, and the adjusted reference confidence coefficient is obtained and used as the confidence coefficient of the other areas after adjustment.

In this way, the first confidence coefficient adjusting value is determined through the intersection ratio between the other area and the first area, the intersection ratio reflects the overlapping ratio of the other area and the first area, the second confidence coefficient adjusting value is determined according to the blocked degree of the other area, the blocked degree reflects the blocked degree of the other area, and the first confidence coefficient adjusting value and the second confidence coefficient adjusting value can both reflect the blocked condition of the other area from different angles. Therefore, when the first confidence coefficient adjusting value and the second confidence coefficient adjusting value are adopted to adjust the confidence coefficient of other areas, the first confidence coefficient adjusting value and the second confidence coefficient adjusting value reflect the blocked condition of other areas from different angles, and when the first confidence coefficient adjusting value and the second confidence coefficient adjusting value are adopted to adjust the confidence coefficient based on the blocked condition of other areas which are more accurate, the adjusted confidence coefficient is more approximate to the actual condition.

The above-described loop update confidence is described in one specific implementation.

Assuming that each candidate target region is b1, b2, b3, the preset overlap ratio threshold Nt is 90%, and the confidence and occlusion degrees of each candidate target region are shown in table 1 below.

TABLE 1

Candidate target area	Confidence level	Degree of shielding
			Region b1	Cv1	Co1
Region b2	Cv2	Co2
			Region b3	Cv3	Co3

At the beginning of the first cycle, region set b= { B1, B2, B3}, where confidence Cv1 of region B1 is highest, region B1 is the first region, and regions B2, B3 are other regions.

For the region b2, calculating the intersection ratio between the region b2 and the region b1, and determining a first confidence adjustment value according to the intersection ratio and 90%. And determining a second confidence adjustment value according to the occlusion degree Co2 of the region b 2. And adjusting the confidence coefficient Cv2 of the region b2 according to the first confidence coefficient adjusting value and the second confidence coefficient adjusting value, wherein the updated confidence coefficient is Cv21.

For the region b3, calculating the intersection ratio between the region b3 and the region b1, and determining a first confidence adjustment value according to the intersection ratio and 90%. And determining a second confidence adjustment value according to the occlusion degree Co3 of the region b 3. And adjusting the confidence coefficient Cv3 of the region b3 according to the first confidence coefficient adjusting value and the second confidence coefficient adjusting value, wherein the updated confidence coefficient is Cv31.

The confidence levels for each candidate target region after the first cycle are updated are shown in table 2 below.

TABLE 2

At the beginning of the second cycle, since region B1 has been selected, region set b= { B2, B3}, where confidence Cv21 of region B2 is highest, region B2 is the first region, and region B3 is the other region.

For the region b3, calculating the intersection ratio between the region b3 and the region b2, and determining a first confidence adjustment value according to the intersection ratio and 90%. And determining a second confidence adjustment value according to the occlusion degree Co3 of the region b 3. And adjusting the confidence coefficient Cv31 of the region b3 according to the first confidence coefficient adjusting value and the second confidence coefficient adjusting value, wherein the updated confidence coefficient is Cv311.

Since the regions B1 and B3 have been selected, the region set b= { B3}, including one region, ends the loop.

The confidence of each candidate target region after the final update is shown in table 3 below.

TABLE 3 Table 3

Candidate target area	Confidence level
		Region b1	Cv1
Region b2	Cv21
		Region b3	Cv311

In one embodiment of the present disclosure, in the step S101, target detection may be performed on the image for different target scales, so as to obtain candidate target areas with different scales in the image, confidence levels of the candidate target areas, and blocked degrees of the candidate target areas.

The target scale refers to: the size of the target.

The target scale may be a predetermined scale value, e.g., the target scale may be 16x16, 32x32, 64x64.

Specifically, multiple layers of feature extraction can be performed on the image, and then feature fusion is performed on different features to obtain features with different scales. And carrying out target detection on the image by adopting the features with different scales to obtain candidate target areas with different scales, and obtaining the confidence and the shielding degree of the candidate target areas with different scales.

Therefore, the candidate target areas with different scales contain different image characteristic information, and the characteristic information of the candidate target areas with different scales in the image is enriched by obtaining the candidate target areas with different scales.

In one embodiment of the disclosure, an image may be input into a target detection model trained in advance, and a candidate target region, a confidence level of the candidate target region, and a blocked level of the candidate target region in the image output by the target detection model are obtained.

The object detection model includes: a target detection layer for detecting a candidate target region in the image and an occlusion degree prediction layer for predicting the occlusion degree of the candidate target region.

In one implementation, the target detection layer may calculate the confidence of the candidate target region in addition to detecting the candidate target region in the image, in which case the network structure of the target detection model may be as shown in fig. 3a, and the target detection model includes a target detection layer and a occlusion prediction layer.

Specifically, after the image is input into a target detection model, a target detection layer in the model detects candidate target areas in the image, calculates the confidence coefficient of the candidate target areas, and transmits a detection result to a shielding degree prediction layer; the shielding degree prediction layer predicts the shielding degree of each candidate target area, and the target detection model outputs the candidate target area, the confidence degree of the candidate target area and the shielding degree.

As known from the foregoing embodiments, when the image is subjected to target detection, candidate target areas with different scales, confidence levels of the candidate target areas, and shielding levels of the candidate target areas can be obtained for each target scale. In this case, FPN (Feature Pyramid Networks, feature pyramid network) may be added to the network model, where FPN is used to obtain candidate target regions of various scales, confidence levels of candidate target regions, and occlusion levels.

The network structure of the network model after adding the FPN may be as shown in fig. 3b, and the network structure shown in fig. 3b includes a Backbone network (Backbone) and the FPN.

The backbone network is used for extracting features of the image to obtain image features of different levels in the image, and inputting the image features of different levels into the FPN.

For example: when the backbone network is a convolutional neural network, each layer of convolutional layer of the convolutional neural network can carry out convolutional operation on the image, so that image features of different levels are obtained.

The FPN is used for carrying out feature fusion on the image features of different levels to obtain the image features of different scales, carrying out target detection based on the image features of different scales to obtain candidate target areas of different scales, obtaining the confidence and the blocked degree of the candidate target areas, and realizing the treatment of the image features of different levels.

When the target detection model is trained, the sample image is used as a training sample, the real candidate target area and the real shielding degree in the sample image are used as training marks, and the preset neural network model is trained until the training ending condition is met, so that the trained target detection model is obtained.

The predetermined neural network model may be a CNN (Conv Neural Network, convolutional neural network) model, an RNN (Recurrent Neural Network ) model, a DNN (Deep Neural Network, deep neural network) model, or the like.

Specifically, after the sample image is input into a preset neural network model, the preset neural network model carries out target detection on the sample image to obtain a candidate target area and a shielded degree of the sample image, the difference between the candidate target area and a real target area and the difference between the shielded degree and the real shielded degree of the candidate target area are calculated, and parameters of the neural network model are adjusted according to the calculated difference, so that the parameters are continuously iteratively adjusted until preset training ending conditions are met.

The training ending condition may be that the training times reach a preset number of times, the model parameters meet a preset model parameter convergence condition, and the like.

The target detection model is obtained through training a large number of training samples, and in the training process, the target detection model learns the characteristics of the target region and the blocked characteristics in the sample image, so that the target detection model has strong robustness, and accurate candidate target regions, confidence degrees of the candidate target regions and blocked degrees can be output when the target detection model is adopted to detect the target of the image.

In step S101 described above, in addition to performing object detection on an image using an object detection model, the image may be divided into a plurality of regions, image features in the region may be extracted for each region, and candidate object regions in the region may be determined from the image features.

The image features include: texture features, color features, edge features, etc.

After each candidate target area is obtained, the confidence of each candidate target is predicted according to the image characteristics of each candidate target area.

And, the occlusion degree of each candidate target area may be calculated according to the layer to which each candidate target area belongs and the position information.

Specifically, whether the candidate target areas are shielded or not can be determined according to the relative relation between the layers and the positions to which the candidate target areas belong, and the ratio between the shielded area and the shielded area is calculated and used as the shielded degree of the candidate target areas.

For example: when the candidate target area A is positioned on the foreground layer, the candidate target area B is positioned on the background layer, and the position information of the candidate target area A and the position information of the candidate target area B are overlapped, the occlusion of the candidate target area B can be determined, and the ratio of the occlusion area of the candidate target area B to the occlusion area of the candidate target area B is calculated and is used as the occlusion degree of the candidate target area B.

Corresponding to the method for detecting the target in the vehicle-road cooperation, the embodiment of the disclosure also provides a device for detecting the target in the vehicle-road cooperation.

Referring to fig. 4, fig. 4 is a schematic structural diagram of a target detection device in vehicle-road cooperation according to an embodiment of the disclosure, where the device includes the following modules 401-403.

The information obtaining module 401 is configured to perform target detection on an image to obtain a candidate target area in the image, a confidence level of the candidate target area, and a blocked level of the candidate target area;

a confidence updating module 402, configured to update the confidence of the candidate target region based on the intersection ratio between the candidate target regions and the occlusion degree of the candidate target region;

the target detection module 403 is configured to detect a target in the image from the candidate target area according to the updated confidence.

In one embodiment of the present disclosure, the confidence updating module 402 is specifically configured to circularly select a first region with highest confidence from a region set, update the confidence of other regions according to the intersection ratio of other regions in the region set and the first region and the occlusion degree of other regions until the region set includes one region, where the region set includes: unselected areas in the candidate target area.

In one embodiment of the present disclosure, the confidence update module 402 includes:

the cross-over ratio calculating unit is used for calculating the cross-over ratio between other areas in the area set and the first area;

The first adjustment value determining unit is used for determining a first confidence adjustment value according to the cross ratio and a preset cross ratio threshold;

the second adjustment value determining unit is used for determining a second confidence adjustment value according to the occluded degree of other areas;

and the confidence coefficient adjusting unit is used for adjusting the confidence coefficient of other areas by adopting the first confidence coefficient adjusting value and the second confidence coefficient adjusting value.

In one embodiment of the disclosure, the first adjustment value determining unit is specifically configured to determine whether the blending ratio is smaller than a preset blending ratio threshold; if yes, determining that the first confidence coefficient adjusting value is 1; if not, determining that the first confidence adjustment value is: 1 and the cross ratio.

In one embodiment of the present disclosure, the second adjustment value determining unit is specifically configured to determine the second confidence adjustment value g (occ _pred) according to the following expression:

g(occ_pred)＝α ^occ_pred

In one embodiment of the disclosure, the confidence level adjusting unit is specifically configured to adjust the confidence level of the other region according to the following expression:

S’＝S*T1*T2

wherein S' represents the confidence of other regions after adjustment, S represents the confidence of other regions before adjustment, T1 represents the first confidence adjustment value, and T2 represents the second confidence adjustment value.

In one embodiment of the present disclosure, the target detection module 403 is specifically configured to select a candidate target area with updated confidence coefficient greater than a preset confidence coefficient threshold, and determine a target in the selected candidate target area as a target in the image; or selecting a preset number of candidate target areas with the highest updated confidence coefficient, and determining the targets in the selected candidate target areas as the targets in the image.

Thus, for candidate target regions with a confidence greater than the preset confidence threshold, the likelihood of including targets in those candidate target regions is higher than the likelihood of including targets in other candidate target regions. Therefore, the targets in the candidate target areas with the confidence degrees larger than the preset confidence degree threshold value are determined to be targets in the image, and the obtained targets are high in accuracy; for a preset number of candidate target areas with the highest confidence, the probability of containing targets in the candidate target areas is higher than the probability of containing targets in other candidate areas. Therefore, the targets in the preset number of candidate target areas with the maximum confidence are determined as the targets in the image, and the accuracy of the obtained targets is high.

In one embodiment of the present disclosure, the information obtaining module 401 is specifically configured to perform target detection on an image according to different target scales, so as to obtain candidate target areas with different scales in the image, confidence levels of the candidate target areas, and occlusion levels of the candidate target areas.

In one embodiment of the present disclosure, the information obtaining module 401 is specifically configured to input an image into a target detection model obtained by training in advance, obtain a candidate target area in the image output by the target detection model, a confidence level of the candidate target area, and an occlusion level of the candidate target area, where the target detection model includes: a target detection layer for detecting a candidate target region in the image and an occlusion degree prediction layer for predicting the occlusion degree of the candidate target region.

In this way, the target detection model is obtained through training a large number of training samples, and in the training process, the target detection model learns the characteristics of the target region and the blocked characteristics in the sample image, so that the target detection model has strong robustness, and accurate candidate target regions, confidence degrees of the candidate target regions and blocked degrees can be output when the target detection model is adopted to carry out target detection on the image.

According to embodiments of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium and a computer program product.

In one embodiment of the present disclosure, there is provided an electronic device including:

at least one processor; and

a memory communicatively coupled to the at least one processor; wherein,,

the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of object detection in any one of the vehicle lane co-ordination described in the method embodiments described above.

In one embodiment of the present disclosure, a non-transitory computer readable storage medium storing computer instructions for causing a computer to perform a method of target detection in any of the aforementioned method embodiments is provided.

In one embodiment of the present disclosure, a computer program product is provided, comprising a computer program which, when executed by a processor, implements a method for target detection in any of the aforementioned method embodiments in a vehicle lane collaboration.

In one embodiment of the disclosure, a roadside device is provided, including the electronic device described above.

In one embodiment of the disclosure, a cloud control platform is provided, including the electronic device.

Fig. 5 illustrates a schematic block diagram of an example electronic device 500 that may be used to implement embodiments of the present disclosure. Electronic devices are intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. The electronic device may also represent various forms of mobile devices, such as personal digital processing, cellular telephones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions, are meant to be exemplary only, and are not meant to limit implementations of the disclosure described and/or claimed herein.

As shown in fig. 5, the apparatus 500 includes a computing unit 501 that can perform various suitable actions and processes according to a computer program stored in a Read Only Memory (ROM) 502 or a computer program loaded from a storage unit 508 into a Random Access Memory (RAM) 503. In the RAM 503, various programs and data required for the operation of the device 500 can also be stored. The computing unit 501, ROM 502, and RAM 503 are connected to each other by a bus 504. An input/output (I/O) interface 505 is also connected to bus 504.

Various components in the device 500 are connected to the I/O interface 505, including: an input unit 506 such as a keyboard, a mouse, etc.; an output unit 507 such as various types of displays, speakers, and the like; a storage unit 508 such as a magnetic disk, an optical disk, or the like; and a communication unit 509 such as a network card, modem, wireless communication transceiver, etc. The communication unit 509 allows the device 500 to exchange information/data with other devices via a computer network such as the internet and/or various telecommunication networks.

The computing unit 501 may be a variety of general and/or special purpose processing components having processing and computing capabilities. Some examples of computing unit 501 include, but are not limited to, a Central Processing Unit (CPU), a Graphics Processing Unit (GPU), various specialized Artificial Intelligence (AI) computing chips, various computing units running machine learning model algorithms, a Digital Signal Processor (DSP), and any suitable processor, controller, microcontroller, etc. The calculation unit 501 performs the respective methods and processes described above, for example, the target detection method in the vehicle-road cooperation. For example, in some embodiments, the method of target detection in vehicle-road collaboration may be implemented as a computer software program tangibly embodied on a machine-readable medium, such as storage unit 508. In some embodiments, part or all of the computer program may be loaded and/or installed onto the device 500 via the ROM 502 and/or the communication unit 509. When the computer program is loaded into the RAM 503 and executed by the computing unit 501, one or more steps of the above-described object detection method in the vehicle-road cooperation may be performed. Alternatively, in other embodiments, the computing unit 501 may be configured to perform the method of target detection in vehicle-road coordination in any other suitable way (e.g. by means of firmware).

Various implementations of the systems and techniques described here above may be implemented in digital electronic circuitry, integrated circuit systems, field Programmable Gate Arrays (FPGAs), application Specific Integrated Circuits (ASICs), application Specific Standard Products (ASSPs), systems On Chip (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and/or combinations thereof. These various embodiments may include: implemented in one or more computer programs, the one or more computer programs may be executed and/or interpreted on a programmable system including at least one programmable processor, which may be a special purpose or general-purpose programmable processor, that may receive data and instructions from, and transmit data and instructions to, a storage system, at least one input device, and at least one output device.

Optionally, the road side device may include, besides an electronic device, a communication component, and the electronic device may be integrally integrated with the communication component or may be separately provided. The electronic device may acquire data, such as pictures and videos, of a perception device (e.g., a roadside camera) for image video processing and data computation. Optionally, the electronic device itself may also have a perceived data acquisition function and a communication function, such as an AI camera, and the electronic device may directly perform image video processing and data calculation based on the acquired perceived data.

Optionally, the cloud control platform performs processing at the cloud, and the electronic device included in the cloud control platform may acquire data of the sensing device (such as a roadside camera), for example, pictures, videos, and so on, so as to perform image video processing and data calculation; the cloud control platform can also be called a vehicle-road collaborative management platform, an edge computing platform, a cloud computing platform, a central system, a cloud server and the like.

Program code for carrying out methods of the present disclosure may be written in any combination of one or more programming languages. These program code may be provided to a processor or controller of a general purpose computer, special purpose computer, or other programmable data processing apparatus such that the program code, when executed by the processor or controller, causes the functions/operations specified in the flowchart and/or block diagram to be implemented. The program code may execute entirely on the machine, partly on the machine, as a stand-alone software package, partly on the machine and partly on a remote machine or entirely on the remote machine or server.

In the context of this disclosure, a machine-readable medium may be a tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. The machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, 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 disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

To provide for interaction with a user, the systems and techniques described here can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to a user; and a keyboard and pointing device (e.g., a mouse or trackball) by which a user can provide input to the computer. Other kinds of devices may also be used to provide for interaction with a user; for example, feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form, including acoustic input, speech input, or tactile input.

The systems and techniques described here can be implemented in a computing system that includes a background component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front-end component (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such background, middleware, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: local Area Networks (LANs), wide Area Networks (WANs), and the internet.

The computer system may include a client and a server. The client and server are typically remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other.

It should be appreciated that various forms of the flows shown above may be used to reorder, add, or delete steps. For example, the steps recited in the present disclosure may be performed in parallel or sequentially or in a different order, provided that the desired results of the technical solutions of the present disclosure are achieved, and are not limited herein.

The above detailed description should not be taken as limiting the scope of the present disclosure. It will be apparent to those skilled in the art that various modifications, combinations, sub-combinations and alternatives are possible, depending on design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present disclosure are intended to be included within the scope of the present disclosure.

Claims

1. A method for detecting a target in a vehicle-road collaboration, the method comprising:

performing target detection on an image to obtain a candidate target area in the image, and confidence coefficient and blocked degree of the candidate target area, wherein the candidate target area refers to an area which is considered to possibly exist a target through target detection;

And circularly selecting a first region with highest confidence from the region set, and updating the confidence of other regions according to the intersection ratio of other regions in the region set and the first region and the shielding degree of other regions until the region set comprises one region, wherein the region set comprises: unselected areas in the candidate target area;

2. The method of claim 1, wherein updating the confidence of the other regions according to the intersection ratio of the other regions in the region set and the first region and the occlusion degree of the other regions comprises:

calculating the cross-over ratio between other areas in the area set and the first area;

determining a first confidence coefficient adjusting value according to the cross-over ratio and a preset cross-over ratio threshold;

determining a second confidence coefficient adjusting value according to the blocked degree of other areas;

and adjusting the confidence coefficient of other areas by adopting the first confidence coefficient adjusting value and the second confidence coefficient adjusting value.

3. The method of claim 2, wherein the determining a first confidence adjustment value based on the cross-over ratio and a preset cross-over ratio threshold comprises:

Judging whether the cross ratio is smaller than a preset cross ratio threshold value or not;

if yes, determining that the first confidence coefficient adjusting value is 1;

if not, determining that the first confidence adjustment value is: 1 and the cross ratio.

4. The method of claim 2, wherein the determining a second confidence adjustment value based on the occlusion of the other region comprises:

the second confidence adjustment value g (occ _pred) is determined according to the following expression:

g(occ_pred))＝α ^occ_pred

5. The method of any of claims 2-4, wherein said adjusting the confidence of other regions using the first and second confidence adjustment values comprises:

confidence in other regions is adjusted according to the following expression:

S’＝S*T1*T2

6. The method of any of claims 1-4, wherein the detecting a target in the image from a candidate target region according to the updated confidence comprises:

Selecting a candidate target area with updated confidence coefficient larger than a preset confidence coefficient threshold value, and determining a target in the selected candidate target area as a target in the image;

or (b)

And selecting a preset number of candidate target areas with the highest updated confidence coefficient, and determining the targets in the selected candidate target areas as the targets in the image.

7. The method of any of claims 1-4, wherein the performing object detection on the image to obtain a candidate object region in the image, a confidence level of the candidate object region, and a occlusion level of the candidate object region, comprises:

and aiming at different target scales, carrying out target detection on the image to obtain candidate target areas with different scales in the image, the confidence level of the candidate target areas and the shielding level of the candidate target areas.

8. The method of any of claims 1-4, wherein the performing object detection on the image to obtain a candidate object region in the image, a confidence level of the candidate object region, and a occlusion level of the candidate object region, comprises:

inputting an image into a target detection model which is obtained through training in advance, and obtaining a candidate target area, a confidence level of the candidate target area and a blocked level of the candidate target area in the image which is output by the target detection model, wherein the target detection model comprises: a target detection layer for detecting a candidate target region in the image and an occlusion degree prediction layer for predicting the occlusion degree of the candidate target region.

9. An in-vehicle-road-coordination target detection device, the device comprising:

the information acquisition module is used for carrying out target detection on the image to obtain a candidate target area in the image, the confidence level of the candidate target area and the shielding level of the candidate target area, wherein the candidate target area refers to an area which is considered to possibly exist a target through target detection;

the confidence updating module is used for circularly selecting a first region with highest confidence from the region set, updating the confidence of other regions according to the intersection ratio of other regions in the region set and the first region and the occlusion degree of other regions until the region set comprises one region, wherein the region set comprises: unselected areas in the candidate target area;

10. The apparatus of claim 9, wherein the confidence update module comprises:

11. The apparatus according to claim 10, wherein the first adjustment value determining unit is specifically configured to determine whether the overlap ratio is smaller than a preset overlap ratio threshold; if yes, determining that the first confidence coefficient adjusting value is 1; if not, determining that the first confidence adjustment value is: 1 and the cross ratio.

12. The apparatus of claim 10, wherein the second adjustment value determining unit is specifically configured to determine the second confidence adjustment value g (occ _pred) according to the following expression:

g(occ_pred))＝α ^occ_pred

13. The apparatus according to any of claims 10-12, wherein the confidence adjustment unit is specifically configured to adjust the confidence of the other regions according to the following expression:

S’＝S*T1*T2

14. The apparatus according to any one of claims 9-12, wherein the target detection module is specifically configured to select candidate target areas with updated confidence levels greater than a preset confidence level threshold, and determine a target in the selected candidate target areas as a target in the image; or selecting a preset number of candidate target areas with the highest updated confidence coefficient, and determining the targets in the selected candidate target areas as the targets in the image.

15. The apparatus according to any one of claims 9-12, wherein the information obtaining module is specifically configured to perform object detection on an image for different object scales, to obtain candidate object regions of different scales in the image, confidence levels of the candidate object regions, and occlusion levels of the candidate object regions.

16. The apparatus according to any one of claims 9-12, wherein the information obtaining module is specifically configured to input an image into a pre-trained target detection model, obtain a candidate target region, a confidence level of the candidate target region, and an occlusion level of the candidate target region in the image output by the target detection model, where the target detection model includes: a target detection layer for detecting a candidate target region in the image and an occlusion degree prediction layer for predicting the occlusion degree of the candidate target region.

17. An electronic device, comprising:

at least one processor; and

a memory communicatively coupled to the at least one processor; wherein,,

the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-8.

18. A non-transitory computer readable storage medium storing computer instructions for causing the computer to perform the method of any one of claims 1-8.

19. A roadside device comprising the electronic device of claim 17.

20. A cloud control platform comprising the electronic device of claim 17.