An object recognition method, device, equipment and medium

Abstract

The application provides a simulation object identification method, device, equipment and medium, wherein the method comprises: acquiring image data of a target region in a current frame, and extracting a target contour region of a target object in the image data; performing correlation analysis on feature data of the target contour region corresponding to the current frame and feature data of a target contour region corresponding to a previous frame to obtain a correlation result corresponding to the current frame; and when the correlation result corresponding to the current frame and correlation results corresponding to a plurality of continuous adjacent frames satisfy a preset condition, determining the target object as a simulation object. The application improves the accuracy of simulation object identification by automatically analyzing and judging the target object of the target region, and can analyze simulation objects in the surrounding environment in real time and effectively and accurately identify the simulation objects.

Description

Technical Field

[0001] This invention relates to the field of computer vision technology, specifically to a method, apparatus, device, and medium for recognizing simulated objects. Background Technology

[0002] Intelligent devices are increasingly integrated into people's work and lives. To ensure the intelligence and security of these devices, they need to be able to correctly identify objects, providing a safety guarantee. Currently, in the simulation object recognition technology of intelligent robots, detection and segmentation algorithms are mostly used. However, the accuracy of simulation object recognition in existing technologies is relatively low, leading to a decline in object detection and recognition performance. This results in image acquisition devices being unable to achieve effective high-precision recognition of simulation objects. Summary of the Invention

[0003] Therefore, the technical problem to be solved by the present invention is to overcome the defect of low accuracy in the recognition of simulated objects in the prior art, thereby providing a method, device, equipment and medium for recognizing simulated objects.

[0004] In a first aspect, the present invention provides a method for identifying simulated objects, comprising the following steps: acquiring image data of a target region in the current frame, extracting the target contour region of the target object in the image data, performing correlation analysis on the feature data of the target contour region corresponding to the current frame and the feature data of the target contour region corresponding to the previous frame to obtain the correlation result corresponding to the current frame; when the correlation result corresponding to the current frame and the correlation results corresponding to multiple consecutive adjacent frames satisfy a preset condition, the target object is identified as a simulated object.

[0005] Optionally, in the simulated object recognition method provided by the present invention, extracting the target contour region of the target object in the image data includes: identifying the target object in the image data; segmenting the image data based on the target object to obtain a segmented region; and extracting the target contour region of the target object in the segmented region.

[0006] Optionally, in the simulated object recognition method provided by the present invention, the association result includes contour movement rate and contour movement direction. When the association result corresponding to the current frame and the association results corresponding to multiple consecutive adjacent frames meet a preset condition, the target object is determined as a simulated object, including: determining a first contour movement rate based on the target contour region corresponding to the current frame and the target contour region corresponding to the previous frame; determining a second contour movement rate based on the target contour region corresponding to the previous frame and the target contour region corresponding to the frame before that; acquiring a first device movement rate when the image acquisition device acquires image data of the current frame and a second device movement rate when acquiring image data of the previous frame. The device moving speed and the trend of the device moving direction are calculated; a first ratio of the first contour moving speed to the first device moving speed and a second ratio of the second contour moving speed to the second device moving speed are calculated; the ratio of the first ratio to the second ratio is determined as a third ratio; based on the contour moving direction corresponding to the current frame and the contour moving directions corresponding to multiple consecutive adjacent frames, the trend of the contour moving direction of the target contour region between multiple adjacent frames is determined; when the third ratio and the trend of the contour moving direction and the trend of the device moving direction both satisfy the preset conditions, the target object is determined to be a simulated object.

[0007] Optionally, in the simulated object recognition method provided by the present invention, the association result includes contour overlap degree and / or contour intersection-union ratio. When the association result corresponding to the current frame and the association results corresponding to multiple consecutive adjacent frames meet a preset condition, the target object is determined as a simulated object, including: if the contour overlap degree corresponding to the current frame and the contour overlap degree corresponding to multiple consecutive adjacent frames are both greater than a predetermined contour overlap degree threshold, and / or the contour intersection-union ratio corresponding to the current frame and the contour intersection-union ratio corresponding to multiple consecutive adjacent frames are both greater than a predetermined contour intersection-union ratio threshold, the target object is determined as a simulated object.

[0008] Optionally, in the simulated object recognition method provided by the present invention, the association result further includes contour feature similarity. When the association result corresponding to the current frame and the association results corresponding to multiple consecutive adjacent frames meet a preset condition, the target object is determined as a simulated object. The method further includes: if the contour feature similarity corresponding to the current frame and the contour feature similarity corresponding to multiple consecutive adjacent frames are both greater than a predetermined contour feature similarity threshold, the target object is determined as a simulated object.

[0009] Optionally, in the simulated object recognition method provided by the present invention, determining the trend of the contour movement direction of the target contour region between multiple adjacent frames based on the contour movement direction corresponding to the current frame and the contour movement directions corresponding to multiple consecutive adjacent frames includes: determining a first contour movement direction based on the target contour region corresponding to the current frame and the target contour region corresponding to the previous frame; determining a second contour movement direction based on the target contour region corresponding to the previous frame and the target contour region corresponding to the next-previous frame; and determining the trend of the contour movement direction of the target contour region between multiple adjacent frames based on the first contour movement direction and the second contour movement direction.

[0010] Optionally, in the simulated object recognition method provided by the present invention, when the third ratio and the trend of the contour movement direction and the trend of the device movement direction both satisfy the preset conditions, the target object is determined to be a simulated object, including: comparing the third ratio with a preset threshold to obtain a third ratio comparison result; performing correlation analysis on the trend of the movement direction of the target contour region and the trend of the movement direction of the device to obtain a movement direction correlation result; and when both the third ratio comparison result and the movement direction correlation result satisfy the preset conditions, the target object is determined to be a simulated object.

[0011] Secondly, the present invention provides a simulated object recognition device, comprising: an image processing module, configured to acquire image data of a target region in the current frame, extract the target contour region of the target object in the image data, and perform correlation analysis on the feature data of the target contour region corresponding to the current frame and the feature data of the target contour region corresponding to the previous frame to obtain the correlation result corresponding to the current frame; and a simulated object judgment module, configured to determine the target object as a simulated object when the correlation result corresponding to the current frame and the correlation results corresponding to multiple consecutive adjacent frames meet preset conditions.

[0012] Thirdly, the present invention provides a computer device comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to perform the simulated object recognition method as provided in the first aspect of the present invention.

[0013] Fourthly, the present invention provides a computer-readable storage medium storing computer instructions for causing the computer to perform the simulated object recognition method provided in the first aspect of the present invention.

[0014] The technical solution of this invention has the following advantages:

[0015] The present invention provides a method, apparatus, device, and medium for identifying simulated objects. It acquires image data of a target object, extracts the target contour region of the target object based on the image data, performs correlation analysis on adjacent target contour regions to obtain correlation results, and determines whether the target object is a simulated object by judging the correlation results of adjacent target contour regions. When the correlation results of the target contour regions meet the conditions for determining a simulated object, the image acquisition device determines that the target object is a simulated object. This invention improves the accuracy of simulated object recognition by automatically analyzing and judging target objects in the target contour region, enabling real-time analysis of simulated objects in the surrounding environment and effective, high-precision identification of simulated objects. Attached Figure Description

[0016] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0017] Figure 1 This is a flowchart illustrating a specific example of the simulated object recognition method in this invention.

[0018] Figure 2 This is a schematic diagram of a specific example of the simulated object recognition device in an embodiment of the present invention;

[0019] Figure 3 This is a schematic diagram of a specific example of a computer device in an embodiment of the present invention. Detailed Implementation

[0020] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] In the description of this invention, it should be noted that the technical features involved in the different embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0022] This embodiment provides a simulated object recognition method, which can be applied to image acquisition devices, such as... Figure 1 As shown, it includes the following steps:

[0023] Step S1: Obtain the image data of the target region in the current frame, extract the target contour region of the target object in the image data, and perform correlation analysis between the feature data of the target contour region corresponding to the current frame and the feature data of the target contour region corresponding to the previous frame to obtain the correlation result corresponding to the current frame.

[0024] In one optional embodiment, the target area can be indoor or outdoor places such as streets, squares, and parks, and the target object includes, but is not limited to, realistic objects such as mannequins, animal models, and sculptures, or objects that are close in proportion to the real human body.

[0025] In one alternative embodiment, the target contour region includes, but is not limited to, the contour edge of the object itself, the circumscribed rectangle, etc.

[0026] In an optional embodiment, the adjacent target contour region is the target contour extracted from the target image of the current frame and the previous frame, or the target contour extracted from the target image of the current frame, the previous frame, and the next previous frame.

[0027] In an optional embodiment, after the target contour region is obtained, the target contour region is added to the time series in chronological order.

[0028] In an optional embodiment, when performing feature association analysis on the target contour region in the contour sequence based on the contour time series, the analyzed feature data includes, but is not limited to, area, direction of increase or decrease, and contour intersection-union ratio, and the analysis methods used include, but are not limited to, convex hull analysis, corner analysis, HOG feature analysis, etc.

[0029] In an optional embodiment, the feature data of adjacent target contour regions include, but are not limited to: contour feature similarity of adjacent target contour regions, contour displacement direction of adjacent target contour regions, contour movement displacement of adjacent target contour regions, contour intersection-to-union ratio of adjacent target contour regions, and contour overlap of adjacent target contour regions. The feature data is stored in the feature storage unit for subsequent use in identifying simulated objects.

[0030] Step S2: If the association result corresponding to the current frame and the association results corresponding to multiple consecutive adjacent frames meet the preset conditions, the target object is determined as the simulation object.

[0031] The simulated object recognition method provided in this invention acquires image data of a target object, extracts the target contour region of the target object based on the image data, and performs correlation analysis on adjacent target contour regions to obtain correlation results. By judging the correlation results of adjacent target contour regions, it determines whether the target object is a simulated object. When the correlation results of the target contour regions meet the conditions for determining a simulated object, the image acquisition device determines that the target object is a simulated object. This invention improves the accuracy of simulated object recognition by automatically analyzing and judging target objects in the target contour region, enabling real-time analysis of simulated objects in the surrounding environment and effective high-precision recognition of simulated objects.

[0032] In an optional embodiment, extracting the target contour region of the target object from the image data includes:

[0033] First, identify the target object in the image data.

[0034] Then, the image data is segmented based on the target object to obtain the segmented regions.

[0035] Deep learning technology is used to detect, recognize, and segment images. Specifically, an algorithm model built with a neural network structure is used to classify image pixels and extract the corresponding category contours. Then, image segmentation algorithms, including but not limited to Mask-R-CNN and DeepLap-V3, are used to segment objects and category regions in the target image. Traditional image processing methods such as image transformation and image encoding compression can also be used here. However, using traditional image processing methods will result in excessive computational load and reduced recognition accuracy when recognizing simulated objects, thus making it impossible to accurately identify simulated objects.

[0036] By using the embodiments of the present invention, an image is segmented into multiple regions, and then the target is processed according to the segmented regions, which can greatly reduce the amount of computation in the image reprocessing process and improve the efficiency of simulated object recognition.

[0037] Finally, the target contour region of the target object is extracted from the segmented region.

[0038] The target contour region of the target object in each image data can be extracted using one or more methods, including but not limited to edge extraction, watershed, contour filling extraction, and minimum bounding rectangle.

[0039] In an optional embodiment, the association result includes the contour movement rate and contour movement direction of adjacent target contour regions, and step S2 specifically includes:

[0040] First, a first contour movement rate is determined based on the target contour region corresponding to the current frame and the target contour region corresponding to the previous frame. Then, a second contour movement rate is determined based on the target contour region corresponding to the previous frame and the target contour region corresponding to the frame before that.

[0041] In an optional embodiment, the contour movement rate can be obtained by dividing the distance the centroid of the target contour region in two adjacent frames or the distance the edge of the target contour region in two adjacent frames moves by the time interval between the two adjacent frames.

[0042] Secondly, the image acquisition device is obtained as follows: first device movement speed when acquiring image data of the current frame, second device movement speed when acquiring image data of the previous frame, and trend of device movement direction.

[0043] In one alternative embodiment, the image acquisition device includes, but is not limited to, intelligent robots, cameras, etc.

[0044] Then, calculate the first ratio of the first contour movement rate to the first device movement rate, and the second ratio of the second contour movement rate to the second device movement rate; determine the ratio of the first ratio to the second ratio as the third ratio.

[0045] Finally, based on the contour movement direction corresponding to the current frame and the contour movement directions corresponding to multiple consecutive adjacent frames, the trend of the contour movement direction of the target contour region between multiple adjacent frames is determined. When the third ratio and the trend of the contour movement direction and the trend of the device movement direction both meet the preset conditions, the target object is determined to be a simulated object.

[0046] For example, when the image acquisition device is an intelligent robot, after acquiring an image of the target area, the intelligent robot obtains the target contour region in the image and performs correlation analysis on the feature data of the target contour region. Since the simulated object is a stationary object and does not have the ability to move, the simulated object moves with the intelligent robot as a reference point. Ideally, the moving speed of the intelligent robot is equal to the moving speed of the target contour. Therefore, the ratio of the moving speed of the device to the moving speed of the target contour can be used to determine whether the target object is a simulated object. Similarly, ideally, the moving direction of the intelligent robot is opposite to the moving direction of the target contour. Therefore, the moving direction of the device and the moving direction of the target contour can be used to determine whether the target object is a simulated object.

[0047] In an optional embodiment, the step of determining the trend of the contour movement direction of the target contour region among multiple adjacent frames based on the contour movement direction corresponding to the current frame and the contour movement directions corresponding to multiple consecutive adjacent frames specifically includes:

[0048] First, the first contour movement direction is determined based on the target contour region corresponding to the current frame and the target contour region corresponding to the previous frame.

[0049] In an optional embodiment, the offset direction of the target contour region corresponding to the current frame relative to the target contour region corresponding to the previous frame is determined as the first contour movement direction.

[0050] Secondly, the second contour movement direction is determined based on the target contour region corresponding to the previous frame and the target contour region corresponding to the frame before that.

[0051] In one optional embodiment, the offset direction of the target contour region corresponding to the previous frame of the current frame relative to the target contour region corresponding to the second previous frame of the current frame is determined as the second contour movement direction.

[0052] Finally, based on the first contour movement direction and the second contour movement direction, the trend of the contour movement direction of the target contour region between multiple adjacent frames is determined.

[0053] In an optional embodiment, the step of determining the target object as a simulated object when the third ratio, the trend of the contour movement direction, and the trend of the device movement direction all satisfy preset conditions specifically includes:

[0054] First, the third ratio is compared with a preset threshold to obtain the third ratio comparison result.

[0055] Secondly, the trend of the movement direction of the target contour area and the trend of the movement direction of the device are correlated to obtain the movement direction correlation results.

[0056] Finally, when both the third ratio judgment result and the movement direction association result meet the preset conditions, the target object is determined as the simulation object.

[0057] In an optional embodiment, if the third ratio is within a preset range, it indicates that the third ratio judgment result meets the condition; if the movement direction association result is reversed, it indicates that the movement direction association result meets the condition, and the target object is determined to be a simulated object. For example, the preset range can be [0.8, 1.2], [0.9, 1.1], etc.

[0058] In an optional embodiment, the association result includes contour overlap, and / or contour intersection-union ratio, and step S2 specifically includes:

[0059] If the contour overlap degree of the current frame is greater than the contour overlap degree of multiple consecutive adjacent frames, and / or the contour intersection-union ratio of the current frame is greater than the contour intersection-union ratio of multiple consecutive adjacent frames, the target object is determined as a simulation object.

[0060] In an optional embodiment, the contour intersection-over-union ratio corresponding to the current frame is the contour intersection-over-union ratio of the target contour region in the current frame and the target contour region in the previous frame. The intersection-over-union ratio of the target contour region is calculated through the following steps:

[0061] Add the pixel values ​​of the target contour region corresponding to the current frame and the target contour region corresponding to the previous frame, and then subtract the unified pixel value of the region. The resulting region is the intersection region of the target contour region corresponding to the current frame and the target contour region corresponding to the previous frame. Add the pixel values ​​of the target contour region corresponding to the current frame and the target contour region corresponding to the previous frame, and then subtract the intersection region. The resulting region is the union region of the target contour region corresponding to the current frame and the target contour region corresponding to the previous frame. Then divide the intersection region by the union region to obtain the contour intersection-union ratio.

[0062] For example, when the threshold for the predetermined overlap is 0.9 and the threshold for the predetermined intersection-union ratio is 0.9, when the intelligent robot determines that the overlap of the current frame and the previous frame of the target contour region is greater than 0.9, and / or the intersection-union ratio of the current frame and the previous frame of the target contour region is greater than 0.9, the target is determined to be a simulated object. If only one of the overlap or intersection-union ratio of the current frame and the previous frame of the target contour region satisfies the corresponding threshold, the target object can be determined to be a simulated object.

[0063] In an optional embodiment, the association result further includes contour feature similarity. In step S2 above, while determining the target object by contour movement rate or contour movement displacement and contour movement direction, or by contour overlap and contour intersection-union ratio, the following is also included:

[0064] If the contour feature similarity of the current frame and the contour feature similarity of multiple consecutive adjacent frames are all greater than the predetermined contour feature similarity threshold, the target object is determined as a simulated object.

[0065] In this embodiment of the invention, it is not possible to determine whether a target object is a simulated object solely based on contour feature similarity. It is necessary to combine at least one of the following conditions to determine whether a target object within the target contour region is a simulated object: contour overlap, contour intersection-to-union ratio, contour movement rate change rate, and the trend of the target contour region's movement direction.

[0066] For example, when the threshold for contour feature similarity is 0.9, the intelligent robot determines the contour feature similarity between the current frame and the previous frame of the target contour region. When the contour feature similarity is greater than 0.9 and the contour overlap is greater than the threshold, the target contour region is determined to be a simulated object. When the contour feature similarity is less than or equal to 0.9, the target contour region is determined to be a non-simulated object.

[0067] In an optional embodiment, one of the following conditions can be selected as the criterion for determining whether the target object is a simulated object:

[0068] 1. The third ratio between the contour movement rate of the target contour region and the device movement rate of the image acquisition device is within a preset range, and the trend of the contour movement direction of the target contour region is opposite to the trend of the movement direction of the image acquisition device.

[0069] 2. The contour overlap of the target contour region is greater than the contour overlap threshold, and / or the contour intersection-union ratio of the target contour region is greater than the contour intersection-union ratio threshold.

[0070] 3. The third ratio between the contour movement rate of the target contour region and the device movement rate of the image acquisition device is within a preset range, and the trend of the contour movement direction of the target contour region is opposite to the trend of the movement direction of the image acquisition device, and the contour overlap of the target contour region is greater than the contour overlap threshold, and / or the contour intersection ratio of the target contour region is greater than the contour intersection ratio threshold.

[0071] 4. The third ratio between the contour movement rate of the target contour region and the device movement rate of the image acquisition device is within a preset range, and the trend of the contour movement direction of the target contour region is opposite to the trend of the movement direction of the image acquisition device, and the contour feature similarity is greater than the contour feature similarity threshold.

[0072] 5. The contour overlap of the target contour region is greater than the contour overlap threshold, and / or the contour intersection-union ratio of the target contour region is greater than the contour intersection-union ratio threshold, and the contour feature similarity is greater than the contour feature similarity threshold.

[0073] 6. The third ratio between the contour movement rate of the target contour region and the device movement rate of the image acquisition device is within a preset range, and the trend of the contour movement direction of the target contour region is opposite to the trend of the movement direction of the image acquisition device, and the contour overlap of the target contour region is greater than the contour overlap threshold, and / or the contour intersection-union ratio of the target contour region is greater than the contour intersection-union ratio threshold, and the contour feature similarity is greater than the contour feature similarity threshold.

[0074] The simulated object recognition method provided in the above embodiments can be applied to image acquisition devices, including but not limited to intelligent robots. When applied to intelligent robots, it enables the intelligent robot to analyze false targets in the surrounding environment in real time and perform effective high-precision recognition of simulated objects.

[0075] This embodiment provides a simulated object recognition device, such as Figure 2 As shown, it includes:

[0076] Image processing module 21 is used to acquire image data of the target region in the current frame, extract the target contour region of the target object in the image data, and perform correlation analysis between the feature data of the target contour region corresponding to the current frame and the feature data of the target contour region corresponding to the previous frame to obtain the correlation result corresponding to the current frame. The detailed description of step S1 in the above embodiment will not be repeated here.

[0077] The simulation object judgment module 22 determines the target object as a simulation object when the association result corresponding to the current frame and the association results corresponding to multiple consecutive adjacent frames meet the preset conditions. The detailed description of step S2 in the above embodiment will not be repeated here.

[0078] This invention provides a computer device, such as... Figure 3 As shown, it includes: at least one processor 31, such as a CPU (Central Processing Unit), at least one communication interface 33, a memory 34, and at least one communication bus 32. The communication bus 32 is used to enable communication between these components. The communication interface 33 may include a display screen or a keyboard; optionally, the communication interface 33 may also include a standard wired interface or a wireless interface.

[0079] The memory 34 can be a high-speed RAM (Ramdom Access Memory) or a non-volatile memory, such as at least one disk storage device. Optionally, the memory 34 can also be at least one storage device located remotely from the processor 31. The processor 31 can execute the simulated object recognition method provided in the above embodiments. The memory 34 stores a set of program code, and the processor 31 calls the program code stored in the memory 34 to execute the simulated object recognition method provided in the above embodiments. The communication bus 32 can be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The communication bus 32 can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 3The term 34 may be represented by a single line, but does not necessarily indicate a single bus or a single type of bus. The memory 34 may include volatile memory, such as random-access memory (RAM); it may also include non-volatile memory, such as flash memory, hard disk drive (HDD), or solid-state drive (SSD); or it may include a combination of the above types of memory. The processor 31 may be a central processing unit (CPU), a network processor (NP), or a combination of a CPU and an NP. The processor 31 may further include a hardware chip. This hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof.

[0080] The aforementioned PLD can be a complex programmable logic device (CPLD), a field-programmable gate array (FPGA), a generic array logic (GAL), or any combination thereof.

[0081] This invention provides a computer-readable storage medium, which also stores program instructions. A processor can invoke the program instructions to implement the simulated object recognition method provided in the above embodiments, as described in this application. This invention also provides a computer-readable storage medium storing computer-executable instructions that can execute the simulated object recognition method provided in the above embodiments. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk drive (HDD), or solid-state drive (SSD), etc.; the storage medium may also include combinations of the above types of memory.

[0082] Obviously, the above embodiments are merely illustrative examples for clear explanation and are not intended to limit the implementation. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all possible implementations here. However, obvious variations or modifications derived therefrom are still within the scope of protection of this invention.

Claims

1. A method of simulant object identification, the method comprising: The method comprises the following steps: obtaining image data of a target region in a current frame, and extracting a target contour region of a target object in the image data, performing correlation analysis on feature data of the target contour region corresponding to the current frame and feature data of a target contour region corresponding to a previous frame to obtain a correlation result corresponding to the current frame; when the correlation result corresponding to the current frame and correlation results corresponding to a plurality of continuous adjacent frames satisfy a preset condition, determining the target object as a simulation object, wherein the correlation result comprises a contour moving speed and a contour moving direction, and when the correlation result corresponding to the current frame and the correlation results corresponding to the plurality of continuous adjacent frames satisfy the preset condition, determining the target object as the simulation object comprises: determining a first contour moving speed according to the target contour region corresponding to the current frame and the target contour region corresponding to the previous frame, determining a second contour moving speed according to the target contour region corresponding to the previous frame and a target contour region corresponding to a frame before the previous frame, obtaining a first device moving speed of an image acquisition device when collecting the image data of the current frame, a second device moving speed of the image acquisition device when collecting image data of the previous frame, and a trend of a device moving direction, calculating a first ratio of the first contour moving speed to the first device moving speed and a second ratio of the second contour moving speed to the second device moving speed, determining a third ratio of the first ratio to the second ratio, determining a trend of the contour moving direction of the target contour region between the plurality of adjacent frames according to the contour moving direction corresponding to the current frame and the contour moving directions corresponding to the plurality of continuous adjacent frames, and determining the target object as the simulation object when the third ratio and the trend of the contour moving direction satisfy the trend of the device moving direction.

2. The simulated object recognition method of claim 1, wherein, extracting a target contour region of a target object in the image data comprises: identifying the target object in the image data; segmenting the image data based on the target object to obtain a segmented region; extracting the target contour region of the target object in the segmented region.

3. The simulated object recognition method of claim 1, wherein, The correlation result comprises a contour coincidence degree and / or a contour intersection and union ratio, and when the correlation result corresponding to the current frame and the correlation results corresponding to the plurality of continuous adjacent frames satisfy the preset condition, determining the target object as the simulation object comprises: if the contour coincidence degree corresponding to the current frame and the contour coincidence degrees corresponding to the plurality of continuous adjacent frames are all greater than a predetermined contour coincidence degree threshold, and / or the contour intersection and union ratio corresponding to the current frame and the contour intersection and union ratios corresponding to the plurality of continuous adjacent frames are all greater than a predetermined contour intersection and union ratio threshold, determining the target object as the simulation object.

4. The simulation object recognition method according to claim 1 or 3, characterized by, The correlation result further comprises a contour feature similarity, and when the correlation result corresponding to the current frame and the correlation results corresponding to the plurality of continuous adjacent frames satisfy the preset condition, determining the target object as the simulation object further comprises: if the contour feature similarity corresponding to the current frame and the contour feature similarities corresponding to the plurality of continuous adjacent frames are all greater than a predetermined contour feature similarity threshold, determining the target object as the simulation object.

5. The simulated object recognition method of claim 1, wherein, According to the contour moving direction corresponding to the current frame and the contour moving direction corresponding to the plurality of adjacent frames, a trend of the contour moving direction of the target contour region between the plurality of adjacent frames is determined, including: According to the target contour region corresponding to the current frame and the target contour region corresponding to the previous frame, a first contour moving direction is determined; According to the target contour region corresponding to the previous frame and the target contour region corresponding to the second previous frame, a second contour moving direction is determined; According to the first contour moving direction and the second contour moving direction, a trend of the contour moving direction of the target contour region between the plurality of adjacent frames is determined.

6. The simulation object recognition method according to claim 1 or 5, wherein When the third ratio and the trend of the contour moving direction and the trend of the device moving direction all satisfy the preset condition, the target object is determined as a simulation object, including: The third ratio is compared with a preset threshold to obtain a third ratio comparison result; The trend of the moving direction of the target contour region and the trend of the moving direction of the device are associated and analyzed to obtain a moving direction association result; When the third ratio comparison result and the moving direction association result both satisfy the preset condition, the target object is determined as a simulation object.

7. An artificial object recognition device, characterized by, Including: An image processing module is configured to acquire image data of a target region in a current frame, extract a target contour region of a target object in the image data, and perform associated analysis on feature data of the target contour region corresponding to the current frame and feature data of the target contour region corresponding to a previous frame to obtain an associated result corresponding to the current frame. A simulation object judgment module is configured to determine the target object as a simulation object when the associated result corresponding to the current frame and associated results corresponding to a plurality of consecutive adjacent frames satisfy a preset condition, wherein the associated result includes a contour moving rate and a contour moving direction. The simulation object judgment module is further configured to: determine a first contour moving rate according to the target contour region corresponding to the current frame and the target contour region corresponding to the previous frame, determine a second contour moving rate according to the target contour region corresponding to the previous frame and the target contour region corresponding to a second previous frame, acquire a first device moving rate of an image acquisition device when acquiring image data of the current frame, a second device moving rate of the image acquisition device when acquiring image data of the previous frame, and a trend of a device moving direction, calculate a first ratio of the first contour moving rate and the first device moving rate and a second ratio of the second contour moving rate and the second device moving rate, determine a third ratio as a ratio of the first ratio and the second ratio, determine a trend of the contour moving direction of the target contour region between the plurality of adjacent frames according to the contour moving direction corresponding to the current frame and the contour moving direction corresponding to the plurality of adjacent frames, and determine the target object as a simulation object when the third ratio and the trend of the contour moving direction and the trend of the device moving direction all satisfy the preset condition.

8. A computer device, comprising: Including: At least one processor; and a memory connected with the at least one processor in communication; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to perform the method for identifying a simulated object according to any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, The computer readable storage medium stores computer instructions for causing the computer to perform the method for identifying a simulated object according to any one of claims 1-6.

Images