License Plate Detection

Routine Experiment Effect

In this section, we will learn how to use K230 to implement the license plate detection function. License plate detection can be used in many scenarios, especially when combined with license plate recognition, which we will learn in the next section. We connect to the IDE and run the routine to identify the license plate.

【Original image】

【Detection】

The current routine has added serial port output

For the protocol format, please refer to [Routine Communication Protocol.xlsx] in the document.

Code Explanation

Code structure

Initialization phase:

• Configure basic parameters
• Initializing Pipeline
• Loading the model
• Setting up the AI2D preprocessor

Main loop phase:

• Image acquisition and preprocessing
• Model inference and post-processing
• Results
• Resource Recycling

Exit Processing:

• Abnormality Check
• Resource Cleanup

Code Analysis

For the complete code, please refer to the file [Source Code/09.Scene/06.licence_det.py]

xxxxxxxxxx
# 车牌检测类 License Plate Detection Class
class LicenceDetectionApp(AIBase):
    """
    车牌检测应用类，继承自AIBase
    License plate detection application class, inherited from AIBase
    """
    def __init__(self, kmodel_path, model_input_size, confidence_threshold=0.5, nms_threshold=0.2, rgb888p_size=[224,224], display_size=[1920,1080], debug_mode=0):
        """
        初始化函数 Initialization function
        参数 Parameters:
            kmodel_path: 模型路径 Model path
            model_input_size: 模型输入尺寸 Model input size
            confidence_threshold: 置信度阈值 Confidence threshold
            nms_threshold: NMS阈值 NMS threshold
            rgb888p_size: 输入图像尺寸 Input image size
            display_size: 显示尺寸 Display size
            debug_mode: 调试模式 Debug mode
        """
        super().__init__(kmodel_path, model_input_size, rgb888p_size, debug_mode)
        self.kmodel_path = kmodel_path
        self.model_input_size = model_input_size
        self.confidence_threshold = confidence_threshold
        self.nms_threshold = nms_threshold
        # 确保宽度是16的倍数 Ensure width is multiple of 16
        self.rgb888p_size = [ALIGN_UP(rgb888p_size[0], 16), rgb888p_size[1]]
        self.display_size = [ALIGN_UP(display_size[0], 16), display_size[1]]
        self.debug_mode = debug_mode
        
        # 初始化AI2D实例用于图像预处理 Initialize AI2D instance for image preprocessing
        self.ai2d = Ai2d(debug_mode)
        self.ai2d.set_ai2d_dtype(nn.ai2d_format.NCHW_FMT, nn.ai2d_format.NCHW_FMT, np.uint8, np.uint8)
​
    def config_preprocess(self, input_image_size=None):
        """
        配置图像预处理参数 Configure image preprocessing parameters
        """
        with ScopedTiming("set preprocess config", self.debug_mode > 0):
            ai2d_input_size = input_image_size if input_image_size else self.rgb888p_size
            # 配置双线性插值方法 Configure bilinear interpolation method
            self.ai2d.resize(nn.interp_method.tf_bilinear, nn.interp_mode.half_pixel)
            self.ai2d.build([1,3,ai2d_input_size[1],ai2d_input_size[0]],[1,3,self.model_input_size[1],self.model_input_size[0]])
​
    def postprocess(self, results):
        """
        后处理函数 Postprocessing function
        对模型输出结果进行处理 Process model output results
        """
        with ScopedTiming("postprocess", self.debug_mode > 0):
            det_res = aidemo.licence_det_postprocess(results, 
                                                   [self.rgb888p_size[1], self.rgb888p_size[0]], 
                                                   self.model_input_size, 
                                                   self.confidence_threshold, 
                                                   self.nms_threshold)
            return det_res
​
    def draw_result(self, pl, dets):
        """
        绘制检测结果 Draw detection results
        参数 Parameters:
            pl: PipeLine实例 PipeLine instance
            dets: 检测结果 Detection results
        """
        with ScopedTiming("display_draw", self.debug_mode > 0):
            if dets:
                pl.osd_img.clear()
                point_8 = np.zeros((8), dtype=np.int16)
                for det in dets:
                    # 坐标转换 Coordinate conversion
                    for i in range(4):
                        x = det[i * 2 + 0] / self.rgb888p_size[0] * self.display_size[0]
                        y = det[i * 2 + 1] / self.rgb888p_size[1] * self.display_size[1]
                        point_8[i * 2 + 0] = int(x)
                        point_8[i * 2 + 1] = int(y)
                    # 绘制检测框 Draw detection box
                    for i in range(4):
                        pl.osd_img.draw_line(point_8[i * 2 + 0], 
                                           point_8[i * 2 + 1], 
                                           point_8[(i + 1) % 4 * 2 + 0], 
                                           point_8[(i + 1) % 4 * 2 + 1], 
                                           color=(255, 0, 255, 0), 
                                           thickness=4)
            else:
                pl.osd_img.clear()

flow chart