Image recognition method and system, electronic equipment and storage medium

Abstract

The invention provides an image recognition method and system, electronic equipment and a storage medium. The method comprises the steps of collecting a target image data packet at a vehicle returning coordinate of a target vehicle through a fixed shooting angle; compressing the target image data packet according to the header file of the target image data packet and the color information in the target image data packet to obtain a compressed bitmap; according to the compressed bitmap and environment information corresponding to the target image data packet, generating a vehicle returning behavior identification instruction, and sending the vehicle returning behavior identification instruction to a cloud end through a central control end of the target vehicle, and the cloud decodes the compressed bitmap according to the environment information, performs vehicle returning behavior recognition and outputs a vehicle returning behavior recognition result of the target vehicle. According to the method, the image data volume is compressed through self-adaptive transcoding, accurate recognition of the vehicle returning normativity is achieved through decoding, the image recognition quality in the shared vehicle normative vehicle returning scene is guaranteed, meanwhile, the image recognition speed is increased, and then the user experience feeling is improved.

Description

Technical Field

[0001] The present invention relates to the field of intelligent transportation technology, and in particular to an image recognition method, system, electronic equipment and storage medium. Background Art

[0002] In the standardized return scenario of shared vehicles, image recognition technology can automatically identify whether the vehicle is parked in the specified position and direction, thereby ensuring the cleanliness and orderliness of the return area. Therefore, how to achieve efficient and accurate image recognition is an important topic that urgently needs to be studied in this field.

[0003] In the current standardized return scenario of shared vehicles, the camera of the shared vehicle usually takes a photo of the return location and transmits the captured image to the cloud for standardized return identification. However, the amount of image data captured by the camera is usually large, which results in a large amount of bandwidth resources being consumed during the data transmission process, resulting in low efficiency of image recognition.

[0004] Therefore, there is an urgent need to provide an image recognition method, system, electronic device and storage medium to solve the above problems. Summary of the invention

[0005] The present invention provides an image recognition method, system, electronic device and storage medium, which are used to solve the defect in the prior art that a large amount of bandwidth resources need to be consumed during data transmission, resulting in low efficiency of image recognition, and to improve the efficiency of image recognition while ensuring the quality of image recognition.

[0006] The present invention provides an image recognition method, comprising: Collect target image data packets at the return coordinates of the target vehicle by fixing the shooting angle; compressing the target image data packet according to the header file of the target image data packet and the color information in the target image data packet to obtain a compressed bitmap; Generate a vehicle return behavior identification instruction according to the compressed bitmap and the environmental information corresponding to the target image data packet, and send the vehicle return behavior identification instruction to the cloud through the central control terminal of the target vehicle; Among them, the vehicle return behavior identification instruction is used to instruct the cloud to decode the compressed bitmap according to the environmental information and then perform vehicle return behavior identification, and output the vehicle return behavior identification result of the target vehicle.

[0007] According to an image recognition method provided by the present invention, the target image data packet is compressed according to the header file of the target image data packet and the color information in the target image data packet to obtain a compressed bitmap, including: Deleting the image attribute information corresponding to the fixed shooting angle from the header file of the target image data packet to obtain a bitmap to be processed; the image attribute information includes height information, width information and bit depth information; Grayscale transcoding is performed according to the intensity values ​​of different color channels of each pixel in the bitmap to be processed to obtain the compressed bitmap.

[0008] According to an image recognition method provided by the present invention, grayscale transcoding is performed according to the intensity values ​​of different color channels of each pixel in the bitmap to be processed to obtain the compressed bitmap, including: Performing weighted addition on the intensity values ​​of different color channels of each of the pixel points to obtain the grayscale value of each of the pixel points; The grayscale values ​​of all the pixels are combined to obtain the compressed bitmap.

[0009] According to an image recognition method provided by the present invention, the step of sending the vehicle return behavior recognition instruction to the cloud through the central control terminal of the target vehicle includes: The vehicle return behavior identification instruction is transmitted to the central control terminal via a serial bus, so that the central control terminal can transmit the vehicle return behavior identification instruction to the cloud via a mobile communication network.

[0010] The present invention also provides an image recognition method, comprising: Parse the vehicle return behavior recognition instructions transmitted by the camera to obtain the compressed bitmap and environmental information; Decoding the compressed bitmap according to the environmental information to obtain a decoded bitmap; Generate a data packet according to the image attribute information corresponding to the fixed shooting angle of the camera and the decoding bitmap to obtain a decoded image data packet; Performing vehicle return behavior recognition on the decoded image data packet to obtain a vehicle return behavior recognition result of the target vehicle; Among them, the vehicle return behavior identification instruction is generated by the camera collecting the target image data packet at the vehicle return coordinates of the target vehicle through the fixed shooting angle, and compressing the compressed bitmap according to the header file of the target image data packet and the color information in the target image data packet, and then generating the compressed bitmap according to the environmental information corresponding to the compressed bitmap and the target image data packet.

[0011] According to an image recognition method provided by the present invention, decoding the compressed bitmap according to the environmental information to obtain a decoded bitmap includes: Correcting the grayscale value of each pixel in the compressed bitmap according to the environmental information to obtain a corrected grayscale value of each pixel; Decoding is performed according to the corrected grayscale value of each pixel to obtain intensity values ​​of different color channels of each pixel; The intensity values ​​of different color channels of all the pixels are combined according to the position sequence of the pixels to obtain the decoded bitmap.

[0012] According to an image recognition method provided by the present invention, the data packet is generated according to the image attribute information corresponding to the fixed shooting angle of the camera and the decoding bitmap to obtain a decoded image data packet, including: Filling the decoded bitmap with a header file according to the image attribute information to obtain a filled header file; A data packet is generated according to the filled header file and the decoded bitmap to obtain the decoded image data packet.

[0013] The present invention also provides an image recognition system, comprising: a camera, a central control terminal and a cloud; The camera is connected to the central control terminal via a serial bus, and the central control terminal is connected to the cloud via a mobile communication network; The camera is used to perform any of the above-mentioned image recognition methods; The cloud is used to execute any one of the image recognition methods described above.

[0014] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, any of the above-mentioned image recognition methods is implemented.

[0015] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, and the computer program, when executed by a processor, implements any of the above-mentioned image recognition methods.

[0016] The present invention also provides a computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the image recognition method described above is implemented.

[0017] The image recognition method, system, electronic device and storage medium provided by the present invention collect target image data packets through a fixed shooting angle of a camera, perform adaptive transcoding according to header files and color information to compress the image data volume, so as to reduce bandwidth resource consumption during image data transmission, and generate a return vehicle behavior recognition instruction in combination with the compressed bitmap and environmental information, which is sent to the cloud for accurate decoding to achieve accurate recognition of the return vehicle standardization, thereby ensuring the image recognition quality in the shared vehicle standard return scenario while improving the image recognition speed, thereby improving the user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0019] Figure 1 It is one of the flow charts of the image recognition method provided by the present invention.

[0020] Figure 2 This is the second flow chart of the image recognition method provided by the present invention.

[0021] Figure 3 This is the third flow chart of the image recognition method provided by the present invention.

[0022] Figure 4 It is a structural schematic diagram of the image recognition system provided by the present invention.

[0023] Figure 5 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0024] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0025] In the standard vehicle return scenario, current image recognition technology is mainly divided into image recognition using AI recognition algorithms integrated inside the camera. However, due to the computing power bottleneck inside the camera, image recognition failures can easily occur.

[0026] In this regard, some scholars have proposed an image recognition technology that only collects images inside the camera and transmits the collected images directly to the cloud for regular recognition of vehicle returns. The recognition success rate of the image recognition technology here is guaranteed due to the large amount of computing power in the cloud, but the amount of image data collected by the camera is usually large, resulting in a large amount of bandwidth resources being consumed during the data transmission process, which leads to a bottleneck in the efficiency of image recognition.

[0027] In this regard, the present application provides an image recognition method, which compresses the amount of image data through adaptive transcoding when transmitting the image data from the camera to the cloud, so as to reduce the traffic overhead during the image data transmission process, and decodes it when transmitting it to the cloud to achieve accurate recognition of the return vehicle regulations, thereby ensuring the quality of image recognition while improving the speed of image recognition, thereby improving the user experience.

[0028] Figure 1 is one of the flow charts of the image recognition method provided by the present invention; the execution subject of the method is the camera required for the return vehicle normative identification, which can be a vehicle-mounted camera, etc., and this embodiment does not specifically limit this. The target vehicle in this embodiment is the vehicle required for the return vehicle normative identification, which can be a shared electric vehicle, a shared electric motorcycle, etc., and this embodiment does not specifically limit this.

[0029] like Figure 1 As shown, the method includes step 110 , step 120 and step 130 .

[0030] Step 110 , collecting a target image data packet at the return coordinates of the target vehicle by fixing the shooting angle.

[0031] Optionally, when it is determined that the target vehicle is in a return state, the camera can collect image data at the return coordinates of the target vehicle by fixing the shooting angle to obtain a target image data packet with consistent viewing angle and composition. The return coordinates here can be real-time feedback from the central control terminal based on the return state and parking position of the target vehicle.

[0032] Step 120 compresses the target image data packet according to the header file of the target image data packet and the color information in the target image data packet to obtain a compressed bitmap.

[0033] Figure 2 FIG. 2 is a flow chart of the image recognition method provided by the present invention. Figure 2 As shown, after acquiring the target image data packet, the camera can parse the target image data packet to obtain the corresponding header file and color information, and compress (or transcode) the target image data packet based on the header file and color information, thereby obtaining a compressed bitmap, so as to achieve significant reduction in the size of the image data while retaining the key information in the image, thereby reducing bandwidth consumption.

[0034] The color information includes the intensity values ​​of the three color channels: red (R), green (G), and blue (B). The header file includes basic information about the image, including but not limited to image format, size, and bit depth.

[0035] The compression here can be to encode (such as Huffman coding, etc.) and optimize the target image data packet by combining the header file and color information to reduce the data volume of the target image data packet; or to delete data and convert grayscale of the target image data packet to reduce the data volume of the target image data packet. This embodiment does not specifically limit this.

[0036] Step 130, generating a vehicle return behavior identification instruction according to the compressed bitmap and the environmental information corresponding to the target image data packet, and sending the vehicle return behavior identification instruction to the cloud through the central control terminal of the target vehicle; Among them, the vehicle return behavior identification instruction is used to instruct the cloud to decode the compressed bitmap according to the environmental information and then perform vehicle return behavior identification, and output the vehicle return behavior identification result of the target vehicle.

[0037] Optionally, after compressing the compressed bitmap, the camera can generate instructions based on the compressed bitmap and the environmental information corresponding to the target image data packet to obtain the return behavior recognition instruction, so as to transmit the compressed bitmap and the environmental information corresponding to the target image data packet to the cloud in the form of instructions. The instruction generation here can be implemented by directly combining the compressed bitmap and the environmental information corresponding to the target image data packet to generate the corresponding return behavior recognition instruction, or by filling the compressed bitmap and the environmental information corresponding to the target image data packet into the corresponding position of the instruction template to obtain the corresponding return behavior recognition instruction, which is not specifically limited in this embodiment.

[0038] The environmental information corresponding to the target image data packet here can be configured by the user through the color configuration module fixed on the target vehicle, or it can be acquired by the vehicle-mounted sensor in real time, and this embodiment does not specifically limit this. The configuration can be formed by user input, and the so-called user input can be information input through command line interface input, graphical interface input, touch input, drop-down selection input, voice input, gesture input, visual input, brain-computer input, etc. This embodiment does not specifically limit this.

[0039] The environmental information here includes but is not limited to lighting condition information, such as information obtained by dividing the lighting conditions according to daytime or nighttime.

[0040] After the camera obtains the vehicle return behavior recognition instruction, it can send the vehicle return behavior recognition instruction to the cloud through the central control terminal, so that the cloud can decode and restore the compressed bitmap according to the environmental information and then perform vehicle return behavior recognition, thereby outputting the vehicle return behavior recognition result of the target vehicle. The vehicle return behavior recognition result can be coarse classification information including whether the vehicle return behavior is standardized, or detailed classification information including specific behavior types including whether the vehicle return behavior is standardized or not.

[0041] like Figure 2 As shown, exemplarily, the implementation step of decoding here can be that since the grayscale values ​​of the images collected in the same image collection area under different environmental information are different, in order to ensure the accuracy of recognition, the cloud corrects the grayscale values ​​of each pixel in the compressed bitmap according to the environmental information so that the grayscale values ​​corresponding to the same iconic color remain consistent. Subsequently, the cloud decodes the corrected grayscale values ​​of each pixel according to the decoding method corresponding to the grayscale encoding to obtain the intensity values ​​of different color channels of each pixel. Subsequently, the cloud combines the intensity values ​​of different color channels of each pixel to obtain a decoded bitmap, that is, converting the 1-byte grayscale value to obtain three-byte color data and filling them in sequence to obtain a decoded bitmap containing color information; then, the cloud fills the header file in the data header of the decoded bitmap according to the fixed image attribute information method corresponding to the fixed shooting angle to restore the target image data packet to obtain a decoded image data packet containing a complete header file and color information. Subsequently, the cloud identifies the return vehicle mark position based on the decoded image data packet, and compares the identified return vehicle mark position with the return vehicle coordinates to determine the return vehicle behavior recognition result of the target vehicle, thereby realizing accurate recognition of the return vehicle behavior.

[0042] The method provided in this embodiment collects target image data packets through a fixed shooting angle of the camera, and performs adaptive transcoding according to the header file and color information to compress the image data volume, so as to reduce the bandwidth resource consumption during the image data transmission process, and combines the compressed bitmap and environmental information to generate a return vehicle behavior recognition instruction, which is sent to the cloud for accurate decoding to achieve accurate recognition of the return vehicle standardization, thereby ensuring the image recognition quality in the shared vehicle standard return scenario while improving the image recognition speed, thereby improving the user experience.

[0043] In some embodiments, step 120 specifically includes: Deleting the image attribute information corresponding to the fixed shooting angle from the header file of the target image data packet to obtain a bitmap to be processed; the image attribute information includes height information, width information and bit depth information; Grayscale transcoding is performed according to the intensity values ​​of different color channels of each pixel in the bitmap to be processed to obtain the compressed bitmap.

[0044] Optionally, since the width and height information of the image are fixed after the camera is fixed, and information such as the bit depth in the bitmap (BMP) header file also uses a fixed 24-bit color image, the camera can delete the image attribute information corresponding to the fixed shooting angle from the header file of the target image data packet, that is, delete the 54 bytes used to store the image attribute information in the BMP header file of the target image data packet to obtain the bitmap to be processed, thereby reducing the data transmission amount by reducing the size of the target image data packet and reducing the bandwidth requirement for data transmission.

[0045] Subsequently, the camera can obtain the intensity values ​​of different color channels of each pixel in the bitmap to be processed, and perform grayscale transcoding on the intensity values ​​of different color channels of each pixel in the bitmap to be processed to obtain the grayscale value of each pixel, and combine the grayscale values ​​of each pixel to obtain the corresponding compressed bitmap to further reduce the amount of data transmission.

[0046] The grayscale transcoding here may be weighted fusion of intensity values ​​of different color channels, or averaging, etc., which is not specifically limited in this embodiment.

[0047] The following is a specific example to illustrate the technical effect of grayscale transcoding provided by this embodiment.

[0048] For example, with a resolution of 800 For example, a 24-bit BMP image of 600 pixels is represented by 3 bytes, and the image data size is 800. 600 3 = 1440000 bytes, and through grayscale transcoding, each pixel occupying 3 bytes can be converted into one byte of grayscale, which can directly reduce the image data size to 800 600 1=480,000 bytes, which reduces 938Kb of data. Therefore, the amount of data can be greatly reduced through grayscale transcoding.

[0049] The method provided in this embodiment can reduce the transmission time and data traffic by more than 3 times by deleting image attribute information such as height, width and bit depth in the target image data packet header file, and grayscale transcoding the bitmap to be processed. It only uses a simple algorithm without relying on processor performance, thereby optimizing the user experience and reducing the traffic cost by more than 3 times, thereby effectively reducing the complexity and transmission volume of image data, thereby improving the efficiency and accuracy of image processing, and reducing storage and transmission costs.

[0050] In some embodiments, step 120 further includes: Performing weighted addition on the intensity values ​​of different color channels of each of the pixel points to obtain the grayscale value of each of the pixel points; The grayscale values ​​of all the pixels are combined to obtain the compressed bitmap.

[0051] Optionally, when performing grayscale transcoding, the camera may first perform weighted addition on the intensity values ​​of different color channels of each pixel to obtain the grayscale value of each pixel, and combine the grayscale values ​​of all pixels to obtain a corresponding compressed bitmap.

[0052] The weight values ​​of the intensity values ​​of different color channels here can be adaptively set according to the sensitivity corresponding to different colors. For example, the weight value of the green channel G can be set to about 0.587, the weight value of the red channel R can be set to about 0.299, and the weight value of the blue channel B can be set to about 0.114. Accordingly, the grayscale conversion calculation formula corresponding to the grayscale value Gray of each pixel point can be: Gray=0.299R+0.587G+0.114B.

[0053] The method provided in this embodiment converts the color channel intensity value of each pixel into a grayscale value and combines the grayscale values ​​of all pixels, thereby realizing the conversion from a color image to a grayscale image, while retaining the key brightness information of the image, thereby achieving the effect of simplifying and compressing image data.

[0054] In some embodiments, step 130 specifically includes: The vehicle return behavior identification instruction is transmitted to the central control terminal via a serial bus, so that the central control terminal can transmit the vehicle return behavior identification instruction to the cloud via a mobile communication network.

[0055] Optionally, the camera can communicate with the central control terminal through a serial bus, and the central control terminal can communicate with the cloud through a mobile communication network. Accordingly, during the data transmission process, the camera can first transmit the vehicle return behavior identification instruction to the central control terminal through the serial bus, and then the central control terminal transmits the vehicle return behavior identification instruction to the cloud communication through the mobile communication network.

[0056] The serial bus here may be a 485 bus, and the mobile communication network may be a 4G mobile communication network or a 5G mobile communication network, etc. This embodiment does not specifically limit this.

[0057] The method provided in this embodiment uses a serial bus to transmit the vehicle return behavior recognition instruction to the central control terminal through a camera, and then the central control terminal sends the instruction to the cloud via a mobile communication network, thereby achieving efficient and flexible data transmission, thereby improving the real-time performance of data processing and response.

[0058] In some embodiments, this embodiment further provides an image recognition method, which is executed by the cloud.

[0059] Figure 3 FIG. 3 is a flow chart of the image recognition method provided by the present invention. Figure 3 As shown, the image recognition method includes: step 310, step 320, step 330 and step 340.

[0060] Step 310, parsing the vehicle return behavior identification instruction transmitted by the camera to obtain a compressed bitmap and environmental information; wherein the vehicle return behavior identification instruction is generated by the camera collecting a target image data packet at the vehicle return coordinates of the target vehicle through the fixed shooting angle, and compressing the compressed bitmap according to the header file of the target image data packet and the color information in the target image data packet, and then generating the compressed bitmap according to the compressed bitmap and the environmental information corresponding to the target image data packet.

[0061] Optionally, the cloud can communicate with the central control end in real time to receive the vehicle return behavior recognition instruction transmitted by the camera through the central control end. The vehicle return behavior recognition instruction here is generated by the camera collecting the target image data packet at the vehicle return coordinates of the target vehicle through a fixed shooting angle, and compressing the compressed bitmap according to the header file of the target image data packet and the color information in the target image data packet, and then generating the compressed bitmap according to the environmental information corresponding to the compressed bitmap and the target image data packet. The specific implementation process is shown in steps 110-130, which will not be repeated here.

[0062] After the cloud obtains the vehicle return behavior recognition instruction, it can parse the vehicle return behavior recognition instruction to obtain the compressed bitmap and environmental information therefrom.

[0063] Step 320: decode the compressed bitmap according to the environment information to obtain a decoded bitmap.

[0064] like Figure 2 As shown, after the cloud obtains the environmental information and the compressed bitmap, it can decode the compressed bitmap according to the environmental information to obtain a decoded bitmap. For example, the compressed bitmap can be corrected based on the environmental information, and the corrected compressed bitmap can be decoded based on the decoding step corresponding to the compression step to restore the color information and obtain a decoded bitmap. The correction here can be an overall correction of the compressed bitmap, or a pixel-by-pixel correction of the compressed bitmap, which is not specifically limited in this embodiment.

[0065] Step 330 , generating a data packet according to the image attribute information corresponding to the fixed shooting angle of the camera and the decoding bitmap to obtain a decoded image data packet.

[0066] like Figure 2As shown, after the cloud obtains the decoded bitmap, it can generate a data packet based on the image attribute information and the decoded map corresponding to the fixed shooting angle of the camera to restore the target image data packet, thereby obtaining a decoded image data packet containing complete image attribute information and color information.

[0067] The data packet generation method here includes filling the image attribute information and the decoding map into the corresponding positions of the data packet template of the target image data packet to obtain a decoded image data packet; or, integrating the image attribute information and the decoding map according to the integration format of the target image data packet to obtain a decoded image data packet.

[0068] Step 340: Perform vehicle return behavior recognition on the decoded image data packet to obtain a vehicle return behavior recognition result of the target vehicle.

[0069] like Figure 2 As shown, after the cloud obtains the decoded image data packet, it can identify the position of the return vehicle mark based on the decoded image data packet, and compare the identified return vehicle mark position with the return vehicle coordinates to determine the return vehicle behavior recognition result of the target vehicle, thereby achieving accurate recognition of the return vehicle behavior. The recognition here can be achieved through an artificial intelligence (AI) algorithm.

[0070] The vehicle return behavior identification result may be coarse classification information including whether the vehicle return behavior is standardized, or fine classification information including specific behavior types including whether the vehicle return behavior is standardized or irregular.

[0071] The method provided in this embodiment can effectively reduce the consumption of bandwidth resources in the process of image data transmission by receiving the compressed bitmap compressed by the camera according to the header file and color information through adaptive transcoding in real time on the cloud, and accurately decode the compressed bitmap according to the environmental information after receiving it, so as to realize accurate recognition of the standardization of the return of the vehicle, thereby ensuring the image recognition quality in the standard return scenario of the shared vehicle, improving the speed of image recognition, and thus improving the user experience.

[0072] In some embodiments, step 320 specifically includes: Correcting the grayscale value of each pixel in the compressed bitmap according to the environmental information to obtain a corrected grayscale value of each pixel; Decoding is performed according to the corrected grayscale value of each pixel to obtain intensity values ​​of different color channels of each pixel; The intensity values ​​of different color channels of all the pixels are combined according to the position sequence of the pixels to obtain the decoded bitmap.

[0073] Optionally, since the grayscale values ​​of images collected in the same image collection area under different environmental information are different, in order to ensure the accuracy of recognition, the cloud corrects the grayscale values ​​of each pixel in the compressed bitmap according to the environmental information so that the grayscale values ​​corresponding to the same iconic color remain consistent. Subsequently, the cloud decodes the corrected grayscale values ​​of each pixel according to the decoding method corresponding to the grayscale encoding to obtain the intensity values ​​of different color channels of each pixel. Subsequently, the cloud combines the intensity values ​​of different color channels of each pixel according to the position order of each pixel to obtain a decoded bitmap, that is, converting the 1-byte grayscale value to obtain three-byte color data and filling them in sequence to obtain a decoded bitmap containing color information.

[0074] The method provided in this embodiment converts the grayscale value into color data and obtains a decoded bitmap containing color information by correcting the grayscale value of each pixel in the compressed bitmap according to the environmental information, decoding to obtain the intensity value of different color channels of each pixel, and then combining them in order of position, thereby ensuring the accuracy of image recognition under different environmental information.

[0075] In some embodiments, step 330 specifically includes: Filling the decoded bitmap with a header file according to the image attribute information to obtain a filled header file; A data packet is generated according to the filled header file and the decoded bitmap to obtain the decoded image data packet.

[0076] Optionally, the cloud can fill the data header of the decoded bitmap with a header file according to the fixed image attribute information corresponding to the fixed shooting angle, and generate a data packet with the filled header file and the decoded bitmap to restore the target image data packet to obtain a decoded image data packet containing a complete header file and color information.

[0077] The method provided in this embodiment fills the header file of the decoded bitmap according to the image attribute information in the cloud, and then Figure 1 The data packet is generated together, thereby successfully restoring the decoded image data packet containing the complete header file and color information.

[0078] The image recognition system provided by the present invention is described below. The image recognition system described below and the image recognition method described above can be referenced to each other.

[0079] Figure 4 is a schematic diagram of the structure of the image recognition system provided by the present invention; Figure 4 As shown, the system includes: a camera, a central control terminal and a cloud; The camera is connected to the central control terminal via a serial bus, and the central control terminal is connected to the cloud via a mobile communication network; The camera is used to perform Figure 1 The image recognition method shown in the figure may be specifically referred to steps 110 - 130 , which will not be described in detail here.

[0080] The cloud is used to execute Figure 3 The image recognition method shown in the figure may be specifically referred to steps 310-340, which will not be described in detail here.

[0081] The system provided in this embodiment can effectively reduce the amount of data in the transmission process to reduce traffic usage and improve recognition efficiency while ensuring the quality of image recognition by transcoding and decoding the original bmp picture data content in a predetermined manner.

[0082] The system provided by the present invention is used to execute the above-mentioned method embodiments. Please refer to the above-mentioned embodiments for the specific processes and detailed contents, which will not be repeated here.

[0083] Figure 5 An example of a physical structure diagram of an electronic device is shown in FIG. Figure 5As shown, the electronic device may include: a processor 510, a communication interface 520, a memory 530 and a communication bus 540, wherein the processor 510, the communication interface 520 and the memory 530 communicate with each other through the communication bus 540. The processor 510 may call the logic instructions in the memory 530 to execute the image recognition method, which includes: collecting a target image data packet at the return coordinates of the target vehicle through a fixed shooting angle; compressing the target image data packet according to the header file of the target image data packet and the color information in the target image data packet to obtain a compressed bitmap; generating a return behavior recognition instruction according to the compressed bitmap and the environmental information corresponding to the target image data packet, and sending the return behavior recognition instruction to the cloud through the central control terminal of the target vehicle; wherein the return behavior recognition instruction is used to instruct the cloud to perform return behavior recognition after decoding the compressed bitmap according to the environmental information, and output the return behavior recognition result of the target vehicle. Alternatively, the vehicle return behavior identification instruction transmitted by the camera is parsed to obtain a compressed bitmap and environmental information; the compressed bitmap is decoded according to the environmental information to obtain a decoded bitmap; a data packet is generated according to the image attribute information corresponding to the fixed shooting angle of the camera and the decoded bitmap to obtain a decoded image data packet; the vehicle return behavior is identified on the decoded image data packet to obtain a vehicle return behavior identification result of the target vehicle; wherein the vehicle return behavior identification instruction is generated by the camera collecting a target image data packet at the vehicle return coordinates of the target vehicle through the fixed shooting angle, and compressing the compressed bitmap according to the header file of the target image data packet and the color information in the target image data packet to obtain the compressed bitmap, and then according to the compressed bitmap and the environmental information corresponding to the target image data packet.

[0084] In addition, the logic instructions in the above-mentioned memory 530 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0085] On the other hand, the present invention also provides a computer program product, which includes a computer program, which can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the image recognition method provided by the above-mentioned methods, which includes: collecting a target image data packet at the return coordinates of the target vehicle by fixing a shooting angle; compressing the target image data packet according to the header file of the target image data packet and the color information in the target image data packet to obtain a compressed bitmap; generating a return behavior identification instruction according to the compressed bitmap and the environmental information corresponding to the target image data packet, and sending the return behavior identification instruction to the cloud through the central control terminal of the target vehicle; wherein the return behavior identification instruction is used to instruct the cloud to perform return behavior identification after decoding the compressed bitmap according to the environmental information, and output the return behavior identification result of the target vehicle. Alternatively, the vehicle return behavior identification instruction transmitted by the camera is parsed to obtain a compressed bitmap and environmental information; the compressed bitmap is decoded according to the environmental information to obtain a decoded bitmap; a data packet is generated according to the image attribute information corresponding to the fixed shooting angle of the camera and the decoded bitmap to obtain a decoded image data packet; the vehicle return behavior is identified on the decoded image data packet to obtain a vehicle return behavior identification result of the target vehicle; wherein the vehicle return behavior identification instruction is generated by the camera collecting a target image data packet at the vehicle return coordinates of the target vehicle through the fixed shooting angle, and compressing the compressed bitmap according to the header file of the target image data packet and the color information in the target image data packet to obtain the compressed bitmap, and then according to the compressed bitmap and the environmental information corresponding to the target image data packet.

[0086] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to execute the image recognition method provided by the above-mentioned methods, the method comprising: collecting a target image data packet at the return coordinates of the target vehicle by a fixed shooting angle; compressing the target image data packet according to the header file of the target image data packet and the color information in the target image data packet to obtain a compressed bitmap; generating a return behavior identification instruction according to the compressed bitmap and the environmental information corresponding to the target image data packet, and sending the return behavior identification instruction to the cloud through the central control terminal of the target vehicle; wherein the return behavior identification instruction is used to instruct the cloud to perform return behavior identification after decoding the compressed bitmap according to the environmental information, and output the return behavior identification result of the target vehicle. Alternatively, the vehicle return behavior identification instruction transmitted by the camera is parsed to obtain a compressed bitmap and environmental information; the compressed bitmap is decoded according to the environmental information to obtain a decoded bitmap; a data packet is generated according to the image attribute information corresponding to the fixed shooting angle of the camera and the decoded bitmap to obtain a decoded image data packet; the vehicle return behavior is identified on the decoded image data packet to obtain a vehicle return behavior identification result of the target vehicle; wherein the vehicle return behavior identification instruction is generated by the camera collecting a target image data packet at the vehicle return coordinates of the target vehicle through the fixed shooting angle, and compressing the compressed bitmap according to the header file of the target image data packet and the color information in the target image data packet to obtain the compressed bitmap, and then according to the compressed bitmap and the environmental information corresponding to the target image data packet.

[0087] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0088] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0089] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An image recognition method, characterized in that: include: Collect target image data packets at the return coordinates of the target vehicle by fixing the shooting angle; compressing the target image data packet according to the header file of the target image data packet and the color information in the target image data packet to obtain a compressed bitmap; Generate a vehicle return behavior identification instruction according to the compressed bitmap and the environmental information corresponding to the target image data packet, and send the vehicle return behavior identification instruction to the cloud through the central control terminal of the target vehicle; Among them, the vehicle return behavior identification instruction is used to instruct the cloud to decode the compressed bitmap according to the environmental information and then perform vehicle return behavior identification, and output the vehicle return behavior identification result of the target vehicle.

2. The image recognition method according to claim 1, characterized in that: The step of compressing the target image data packet according to the header file of the target image data packet and the color information in the target image data packet to obtain a compressed bitmap includes: Deleting the image attribute information corresponding to the fixed shooting angle from the header file of the target image data packet to obtain a bitmap to be processed; the image attribute information includes height information, width information and bit depth information; Grayscale transcoding is performed according to the intensity values ​​of different color channels of each pixel in the bitmap to be processed to obtain the compressed bitmap.

3. The image recognition method according to claim 2, characterized in that: The grayscale transcoding is performed according to the intensity values ​​of different color channels of each pixel in the bitmap to be processed to obtain the compressed bitmap, including: Performing weighted addition on the intensity values ​​of different color channels of each of the pixel points to obtain the grayscale value of each of the pixel points; The grayscale values ​​of all the pixels are combined to obtain the compressed bitmap.

4. The image recognition method according to any one of claims 1 to 3, characterized in that: The step of sending the vehicle return behavior identification instruction to the cloud through the central control terminal of the target vehicle includes: The vehicle return behavior identification instruction is transmitted to the central control terminal via a serial bus, so that the central control terminal can transmit the vehicle return behavior identification instruction to the cloud via a mobile communication network.

5. An image recognition method, characterized in that: include: Parse the vehicle return behavior recognition instructions transmitted by the camera to obtain the compressed bitmap and environmental information; Decoding the compressed bitmap according to the environmental information to obtain a decoded bitmap; Generate a data packet according to the image attribute information corresponding to the fixed shooting angle of the camera and the decoding bitmap to obtain a decoded image data packet; Performing vehicle return behavior recognition on the decoded image data packet to obtain a vehicle return behavior recognition result of the target vehicle; Among them, the vehicle return behavior identification instruction is generated by the camera collecting the target image data packet at the vehicle return coordinates of the target vehicle through the fixed shooting angle, and compressing the compressed bitmap according to the header file of the target image data packet and the color information in the target image data packet, and then generating the compressed bitmap according to the environmental information corresponding to the compressed bitmap and the target image data packet.

6. The image recognition method according to claim 5, characterized in that: The step of decoding the compressed bitmap according to the environment information to obtain a decoded bitmap includes: Correcting the grayscale value of each pixel in the compressed bitmap according to the environmental information to obtain a corrected grayscale value of each pixel; Decoding is performed according to the corrected grayscale value of each pixel to obtain intensity values ​​of different color channels of each pixel; The intensity values ​​of different color channels of all the pixels are combined according to the position sequence of the pixels to obtain the decoded bitmap.

7. The image recognition method according to claim 5 or 6, characterized in that: The step of generating a data packet according to the image attribute information corresponding to the fixed shooting angle of the camera and the decoding bitmap to obtain a decoded image data packet includes: Filling the decoded bitmap with a header file according to the image attribute information to obtain a filled header file; A data packet is generated according to the filled header file and the decoded bitmap to obtain the decoded image data packet.

8. An image recognition system, characterized in that: include: Camera, central control terminal and cloud; The camera is connected to the central control terminal via a serial bus, and the central control terminal is connected to the cloud via a mobile communication network; The camera is used to perform the image recognition method according to any one of claims 1 to 4; The cloud is used to execute the image recognition method as described in any one of claims 5 to 7.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, it implements the image recognition method according to any one of claims 1 to 4, or implements the image recognition method according to any one of claims 5 to 7.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, it implements the image recognition method according to any one of claims 1 to 4, or implements the image recognition method according to any one of claims 5 to 7.

Images