introduction This article refers to the address: http:// According to media reports, China’s traffic accidents caused by blind spots in the rearview mirror account for about 30%. Moreover, with the phenomenon that the "test driver's license" does not cool down, the problem of blind spots in the rearview mirror of the car should not be underestimated. The formation of the digital society provides an opportunity for digital real-time monitoring. The digital real-time monitoring of the rear view scene of the car has become a research hotspot to solve the blind spot problem of the rearview mirror. At present, some digital vehicle monitoring systems have appeared on the market. Commonly, there are monitoring systems with split screen display, monitoring systems with seams and stitching, and the 8th generation "satellite" panoramic driving safety system. The monitoring system of the split screen display is only a simple split screen display of the image, and the scene around the vehicle cannot be displayed on the screen in real time; the monitoring system with stitch stitching does not simply superimpose the image, but the image stitching process is formed. In the middle is the car, surrounded by panoramic images, the disadvantage is that there are obvious stitching seams in the four image stitching; the 8th generation "satellite" panoramic driving safety system uses super wide-angle camera, which can better eliminate image stitching The stitching seam at the place forms a panoramic view of the car. Android system has platform openness, and Google's "Open Car Alliance (OAA)" is committed to the seamless connection between cars and Android devices and built-in Android car system directly in the car; DM3730 Integrating the 1GHz ARM Cortex-A8 core and the 800MHz TMS320C64x+ DSP core, the DSP has unparalleled advantages in digital signal processing and is more suitable for image processing. Therefore, the on-board distributed video surveillance system based on Android and DM3730 has broad application prospects. The vehicle distributed video surveillance system integrates the openness of the Android platform, the high performance of ARM+DSP, the scalability of Ethernet and the plug-and-play of USB camera. It has research significance and application value for realizing digital real-time monitoring of automobiles. 1 system's overall design The vehicle distributed video monitoring system consists of four modules: video capture module, video transmission module, video splicing module and video display module. Figure 1 shows the overall design of the system, and Figure 2 shows the hardware interface between the various modules of the system. Figure 1 Schematic diagram of the overall design of the vehicle distributed video surveillance system Figure 2 Block diagram of the hardware interface of the vehicle distributed video surveillance system 1 Video capture module: AM3715 development board through the USB-HOST interface external USB camera, through the Android operating system Java local call interface [3] (JNI) and V4L2 (video 4 linux 2) video drive framework to capture video and display in real time. 2 Video transmission module: Two (or more) AM3715 and DM3730 development boards are connected via Ethernet. The images captured by the USB camera are transmitted to the ARM end of the DM3730 development board in real time using the RTP multicast protocol and a custom synchronization mechanism. 3 Video splicing module: The ARM side of the DM3730 development board runs the embedded Linux operating system (or Android operating system), and the shared memory is mapped on the ARM side and the DSP side through the TI Codec Engine module, so that two (or more) frames are synchronously received. The image can be accessed simultaneously by the ARM and DSP. Expanding the Embedded Computing Vision Library (EMCV) for in-vehicle applications, and porting and optimizing the SURF open source project OpenSURF, the DSP side can stitch two (or more) images in real time, and finally return the stitching result from the shared memory to the ARM end. 4 Video display module: Video display is done through the cross-platform multimedia library SDL (Simple DirectMedia Layer). Among them, the AM3715 development board displays separate USB camera images, and the DM3730 development board displays the stitched images. 2 video capture transmission and display 2.1 Android V4L2 video capture module V4L2 has emerged from the Linux 2.5.x kernel. To enable the UVC driver and the V4L2 programming framework, first check the Android kernel configuration options to generate the video device file /dev/videoX (X indicates the minor device number). The process of video capture of USB camera using V4L2 [7] includes: (1) opening video device files; (2) checking device attributes; (3) setting video format; (4) frame buffer management; (5) loop collecting video ; (6) Turn off the video device. V4L2 is between the application and the hardware device. The application can access the kernel layer data in three ways: direct read/write mode, memory mapping mode and user pointer mode. Direct read/write mode requires constant copying of data in user space and kernel space, and the efficiency is low; the memory mapping method maps the kernel address to the user address space, and the process can directly read and write memory, avoiding copying of data, and has high efficiency; The memory segment of the user pointer mode is assigned by the application itself. The vehicle distributed video surveillance system adopts a highly efficient memory mapping method, and the system call mmap() can map the kernel address to the user address space. 2.2 RTP video transmission module In view of the high-speed transmission capability and good scalability of Ethernet, the on-board distributed video surveillance system transmits images captured by the USB camera between the Android system and the embedded Linux system through RTP multicast. Document [8] describes the process of real-time video transmission using the RTP library JRTPLIB. In order to ensure the synchronization of image transmission between the two AM3715 development boards and the DM3730 development board, the on-board distributed video surveillance system is designed with a synchronous transmission protocol. The protocol is described as follows: (1) sender 1 Each sender waits for the video frame request command 'R' from the receiver, otherwise the transmission operation is not performed. 2 After receiving the frame request command, the sender first sends a video frame transmission start identifier 0xFE to the multicast address to identify the beginning of a frame of video transmission. 3 Transfer the YUY2 format image to the multicast address in turn, transfer m lines each time, transfer n times, and add the RTP packet transmission sequence number in the first byte position of each RTP packet (the sequence number starts from 0 and increases by 1) . Assume that the YUY2 image has a width of width and a height of height. Since an average pixel occupies 2B, the RTP packet data size per transmission is (2m*width+1)B, and the number of transmissions is n=height/m. 4 At the end of the transmission, the video frame transmission end identifier 0xFF is sent to the multicast address to mark the end of one frame of video transmission. (2) Receiver 1 The receiving end sends a frame request command 'R' to the multicast address, then starts the software dongle and is in the blocking wait state. 2 If the software is not fed to the dog at the end of the timer, indicating that the network communication is faulty, resend the frame request command 'R' to the multicast address and restart the software dongle. 3 Receive RTP data packets from each sender in turn, and restore each frame video according to the IP address and RTP packet transmission sequence number until the video frame transmission end identifier 0xFF is received. 2.3 SDL video display module The YUV format is divided into a packed format (Packed Format) and a flat format (Planner Format) in the storage mode. The three components of the packed format Y, U, and V are continuously cross-stored, and the three components of the planar format are separated by Y, U, and V. storage. In the experiment, the image format acquired by the USB camera is the YUY2 format, and the image after the stitching is the YV12 format. The YUY2 format is a packed format, packed in 4:2:2 mode, each pixel retains the Y component, and the sampling rate of the UV component in the horizontal direction is only 1/2 of the Y component, that is, the storage order is [Y0 U0 Y1 V0] [Y2 U2 Y3 V2] ... [Y2n U2n Y2n+1 V2n]. YV12 is a planar format, and the sampling rate of the UV component in the horizontal direction and the vertical direction is 1/2 of the Y component. In particular, in the UV extraction of the YV12 format, the image is first divided into a number of 2 x 2 square matrices, and then a U component and a V component are extracted on each square matrix. For example, for a 6x4 image, the sampling mode of YV12 is as shown in the figure below, and its storage order is [Y0 Y1 Y2 Y3 Y4 Y5 Y6 Y7 Y8 Y9 Y10 Y11 Y12 Y13 Y14 Y15 Y16 Y17 Y18 Y19 Y20 Y21 Y22 Y23] [U0 U1 U2 U3 U4 U5] [V0 V1 V2 V3 V4 V5 V6].
Wood Grain Diffuser Wood Grain Diffuser,Wood Grain Oil Diffuser,Wood Grain Essential Oil Diffuser,Wood Grain Aroma Diffuser Shenzhen Dituo Electronic Co.,Ltd. , https://www.sz-dituo.com
Multifunctional Wood Grain Aroma Diffuser as below: