1. Demo instructions

The main interface of SDK Demo is shown in Figure 1-1. The main functions include license plate image detection, license plate video detection. The specific instructions are as follows:

(1) License plate picture recognition. Click the button to enter the interface as shown in Figure 1-2. Click the icon to enter the image selection interface as shown in Figure 1-3. Selecting the corresponding license plate image, confirm and return to the detection interface, click “Detection” and then the detected license plate number and detection time will be displayed as shown in Figure 1-4. Note: Image detection requires importing images into SD in advance. Please restart the device if the image is not found in the image selection interface after importing the image.

(2) License plate video recognition. As shown in Figure 1-5, open the camera preview interface. The default preview size of the camera is 640X480. Align the camera lens with the moving license plate, and then the corresponding license plate number and detection time are displayed in real time on the top of the interface.

Figure 1-1 Demo main interface

Figure 1-2 Main screen of license plate image detection

Figure 1-3 License plate image selection interface

Figure 1-4 License plate image detection results

Figure 1-5 License plate video detection interface

1.  Algorithm flow description

The Deepcam SDK uses the deepcam.key network authentication method. You need to obtain the key file provided by Deepcam to make sure that you can use it normally. For details, please refer to the official website registration process or contact the staff directly at https://www.deepcam.cn/.

The algorithm description flow chart is shown in Figure 2-1, which is explained as follows:

(1) Click to enter, the main interface will be displayed;

(2) Selecting picture detection. Selecting the picture, then call the picture detection interface to detect, which can help to display the license plate number and time.

(3) Selecting video detection. Inputing 5 consecutive frames of data, which can help to confirm the characteristics of the second frame and the fifth frame data, and also to confirm whether there is a moving vehicle . After that, it can start to do license plate detection, and to obtain the detected license plate and time.

(4) Finally, the detected information will be displayed in the interface.

Figure 2-1 Flow chart of license plate recognition algorithm

DeepCam AI Open Platform: https://ai.deepcam.cn