人工智能-项目实践-智慧交通-OpenMV交通灯图形识别.zip

import sensor import image import time import lcd from pyb import UART, Timer, LED # 数据包字节标识 Pack_Header1 = 0 Pack_Header2 = 1 Pack_MainCmd = 2 Pack_SubCmd1 = 3 Pack_SubCmd2 = 4 Pack_SubCmd3 = 4 Pack_CheckSum = 6 Pack_Ending = 7 # 功能区分字节 OpenMV_Trafficlight = 0x01 OpenMV_Circle = 0x02 OpenMV_Rectangle = 0x03 OpenMV_ColorNumber = 0x04 OpenMV_AllColorNumber = 0x05 sensor.reset() sensor.set_pixformat(sensor.RGB565) sensor.set_framesize(sensor.QVGA) # 320*240 sensor.set_auto_exposure(False) # 关闭自动白平衡 sensor.skip_frames(10) # 刷新设置 # Initialize the lcd screen. lcd.init() # 创建串口对象 uart = UART(3, 115200, 8, None, 1) # LED对象创建 LED_Red = LED(1) LED_Green = LED(2) LED_Blue = LED(3) # 颜色阈值设定 red_traffic = (40, 72, 52, 81, -7, 60) yellow_traffic = (51, 74, -11, 50, -14, 63) green_traffic = (49, 84, -92, -44, -6, 57) # 图形颜色阈值设定 red = (29, 71, 29, 80, 20, 67) green = (29, 60, -62, -16, 4, 61) blue = (9, 52, -6, 52, -77, -26) yellow = (59, 94, -35, 15, 40, 94) purple = (42, 76, 57, 101, -37, 26) cyan = (47, 94, -69, -8, -47, -3) black = (0, 45, -24, 19, -18, 25) # 图形颜色阈值列表 color_list = [red, green, blue] # 串口发送函数 def UART_Send(src, length): for i in range(length): uart.writechar(src[i]) # 置零list所有成员 def Reset_List(src): for i in range(len(src)): src[i] = 0 # 矩形识别 def Rect_Discern(index): sensor.set_framesize(sensor.QQVGA) img = sensor.snapshot().lens_corr(1.8) n = 0 for c in img.find_rects(threshold=28000): area = (c.x(), c.y(), c.w(), c.h()) statistics = img.get_statistics(roi=area) # 像素颜色统计 if(index > 0 and index <= len(color_list)): ranges = color_list[index - 1] if ranges[0] < statistics.l_mode() < ranges[1] and ranges[2] < statistics.a_mode() < ranges[3] and ranges[4] < statistics.b_mode() < ranges[5]: n = n + 1 else: n = n + 1 print("矩形个数 %f" % n) sensor.set_framesize(sensor.QVGA) return n # 圆形识别（只能检测边缘有阴影的） # def Circle_Discern_odl(index): # sensor.set_framesize(sensor.QQVGA) #img = sensor.snapshot().lens_corr(1.8) #n = 0 # for c in img.find_circles(threshold = 5000, x_margin = 20, y_margin = 20, r_margin = 20,r_min = 30, r_max = 300, r_step = 2): #area = (c.x()-c.r(), c.y()-c.r(), 2*c.r(), 2*c.r()) # statistics = img.get_statistics(roi=area)#像素颜色统计 # if (index > 0 and index <= len(color_list)): #ranges = color_list[index - 1] # if ranges[0] < statistics.l_mode() < ranges[1] and ranges[2] < statistics.a_mode() < ranges[3] and ranges[4] < statistics.b_mode() < ranges[5]: #n = n + 1 # else: #n = n + 1 #print("圆形个数 %f" %n) # sensor.set_framesize(sensor.QVGA) # return n # 圆形识别 def Circle_Discern(): sensor.set_brightness(-3) sensor.set_saturation(3) result_final = [0 for i in range(len(color_list))] for m in range(3): result = [0 for i in range(len(color_list))] for j in range(10): img = sensor.snapshot() for i in range(3): number = 0 for b in img.find_blobs([color_list[i]], area_threshold=50, merge=True): n = 0 for l in img.find_line_segments(roi=[b.x() - 5, b.y() - 5, b.w() + 10, b.h() + 10], merge_distance=26, max_theta_diff=26): if (l.length() < 30): n += 1 #img.draw_line(l.line(), color = (255, 0, 0)) #img.draw_rectangle(b.rect(), color = (255, 0 ,0)) if (n > 3): number += 1 if (number > result[i]): result[i] = number for i in range(len(result)): result_final[i] += result[i] for i in range(len(result_final)): avg = result_final[i] / 5 if (avg > int(avg) + 0.8): result_final[i] = int(avg) + 1 else: result_final[i] = int(avg) print("最终结果：", result_final) sensor.set_brightness(0) sensor.set_saturation(0) UART_Send([0x55, 0x02, 0x91, result_final[0], result_final[1], result_final[2], 0x00, 0xBB], 8) # 颜色个数识别 def Color_Discern(index): n = 0 img = sensor.snapshot() if (index > 0 and index <= len(color_list)): blobs = img.find_blobs([color_list[index - 1]], merge=True) n = len(blobs) else: blobs = img.find_blobs(color_list, merge=True) n = len(blobs) print("色块个数 %f" % n) UART_Send([0x55, 0x02, 0x91, n, 0x00, 0x00, 0x00, 0xBB], 8) return n # 颜色类型总数识别 def Color_Type_Discern(): n = 0 img = sensor.snapshot() for i in range(len(color_list)): blobs = img.find_blobs([color_list[i]], merge=True) if (len(blobs) > 0): n = n + 1 print("色块个数：%f" % n) UART_Send([0x55, 0x02, 0x91, n, 0x00, 0x00, 0x00, 0xBB], 8) return n # 获取交通灯颜色 def Get_TrafficLightColor(img): for blob in img.find_blobs([red_traffic, yellow_traffic, green_traffic], roi=([0, 0, 320, 120]), area_threshold=300, merge=True): img.draw_rectangle(blob.rect()) img.draw_cross(blob.cx(), blob.cy()) return blob.code() # 交通灯识别 def Discern_TrafficLight(): sensor.set_brightness(-3) sensor.set_contrast(-3) sensor.set_saturation(3) traffic_light = 0 for i in range(1, 100): img = sensor.snapshot() traffic_light = Get_TrafficLightColor(img) if(traffic_light in [1, 2, 4]): break sensor.skip_frames(tiem=4000) sensor.set_brightness(0) sensor.set_contrast(0) sensor.set_saturation(0) sensor.set_auto_gain(True) sensor.set_auto_whitebal(True) return traffic_light # 二维码识别 def QRCode_Recognition(): for i in range(10): img = sensor.snapshot() for code in img.find_qrcodes(): content = code.payload() return content # 返回二维码识别结果 def Return_QRCode(result): length = len(result) UART_Send([0x55, 0x02, 0x92, 0x01, length], 5) for i in result: uart.writechar(ord(i)) uart.writechar(0xBB) # 发送交通灯识别结果 def Return_TrafficLightResult(result): if(result == 1): result = 1 print("红色") elif(result == 2): result = 3 print("黄色") elif(result == 4): result = 2 print("绿色") else: result = 3 print("默认：黄色") UART_Send([0x55, 0x02, 0x91, result, 0x00, 0x00, 0x00, 0xBB], 8) # 正常的数据 correctDataTable = [0x55, 0x02, 0x00, 0x00, 0x00, 0x00, 0x00, 0xBB] receiveTable = [0, 0, 0, 0, 0, 0, 0, 0] # 检查串口数据 def Check_Uart(): Reset_List(receiveTable) uartDataReceived = False # 检测串口 if (uart.any()): print("data:") # 接收数据头 for i in range(0, 2): receiveTable[i] = uart.readchar() # 判断数据头 if (receiveTable[i] != correctDataTable[i]): break else: # 检查数据头长度 if (i == 1): # 接收数据位 for j in range(2, 7): receiveTable[j] = uart.readchar() # 判断是否有错误数据 if (receiveTable[j] == -1): print("bad data") break else: # 判断数据内容长度 if (j == 6): # 接收包尾 receiveTable[7] = uart.readchar() # 判断包尾 if (receiveTable[7] == corre