• 星瞳AI VISION软件内测!可以离线标注,训练,并生成OpenMV的模型。可以替代edge impulse https://forum.singtown.com/topic/8206
  • 我们只解决官方正版的OpenMV的问题(STM32),其他的分支有很多兼容问题,我们无法解决。
  • 如果有产品硬件故障问题,比如无法开机,论坛很难解决。可以直接找售后维修
  • 发帖子之前,请确认看过所有的视频教程,https://singtown.com/learn/ 和所有的上手教程http://book.openmv.cc/
  • 每一个新的提问,单独发一个新帖子
  • 帖子需要目的,你要做什么?
  • 如果涉及代码,需要报错提示全部代码文本,请注意不要贴代码图片
  • 必看:玩转星瞳论坛了解一下图片上传,代码格式等问题。
  • 为什么在识别的时候帧率一直下降,还一卡一卡的,求解答,谢谢各位!



    • import sensor, image, time , math , pyb
      from pyb import UART
      sensor.reset()
      sensor.set_pixformat(sensor.RGB565)
      sensor.set_framesize(sensor.QVGA)
      sensor.skip_frames(time = 2000)
      sensor.set_auto_gain(False)
      sensor.set_auto_whitebal(False)
      clock = time.clock()
      a1=0
      uart = UART(3, 115200)
      uart.init(115200, bits=8, parity=None, stop=1)
      
      def find_max(blobs):
          max_size=0
          for blob in blobs:
              if blob[2]*blob[3] > max_size:
                  max_blob = blob
                  max_size = blob[2]*blob[3]
          return max_blob
      
      
      red_threshold_01 = ((20, 58, 49, 127, -128, 127));
      white_threshold_01 = ((75, 100, 96, -24, 31, -7));
      pink_threshold_01 = ((30, 100, 5, 73, -42, 32));
      
      
      while(True):
          clock.tick()
          img = sensor.snapshot().lens_corr(1.8)
          blobs = img.find_blobs([red_threshold_01], roi=[0,0,200,200],pixels_threshold=150, area_threshold=500, merge=True, margin=10);
          blobs1 = img.find_blobs([white_threshold_01], pixels_threshold=100, area_threshold=100, merge=True, margin=10);
          if blobs:
          #如果找到了目标颜色
              print("ored")
              for b in blobs:
              #迭代找到的目标颜色区
                  x = b[0]
                  y = b[1]
                  width = b[2]
                  height = b[3]
                      # Draw a rect around the blob.
                  img.draw_rectangle([x,y,width,height]) # rec
                  img.draw_cross(b[5], b[6]) # cx, cy
                  #在目标颜色区域的中心画十字形标记
                  print(b[5],b[6],b[7])
                  X =int(b[5]-img.width()/2)
                  Y =int(b[6]-img.height()/2)
                  if 9<X<100:
                      datax =("ao"+str(X))
                      print("X坐标: ",datax)
                  elif 0<X<10:
                      datax =("aoo"+str(X))
                      print("X坐标: ",datax)
                  elif X==0:
                      datax =("aooo")
                      print("X坐标: ",datax)
                  elif X>99:
                      datax =("a"+str(X))
                      print("X坐标: ",datax)
                  elif -100<X<-9:
                      datax =("b"+str(X))
                      print("X坐标: ",datax)
                  elif -10<X<0:
                      datax =("boo"+str(X))
                      print("X坐标: ",datax)
                  elif X<-99:
                      datax =("b"+str(X))
                      print("X坐标: ",datax)
      
                  if 9<Y<100:
                      datay =("ao"+str(Y))
                      print("Y坐标: ",datay)
                  elif 0<Y<10:
                      datay =("aoo"+str(Y))
                      print("Y坐标: ",datay)
                  elif Y==0:
                      datay =("aooo")
                      print("Y坐标: ",datay)
                  elif Y>99:
                      datay =("a"+str(Y))
                      print("Y坐标: ",datay)
                  elif -100<Y<-9:
                      datay =("b"+str(Y))
                      print("Y坐标: ",datay)
                  elif -10<Y<0:
                      datay =("boo"+str(Y))
                      print("Y坐标: ",datay)
                  elif Y<-99:
                      datay =("b"+str(Y))
                      print("Y坐标: ",datay)
      
      
                  if(a1==60):
                      a1=0
                      uart.write(datax+datay+"\r\n")
                      print("X Y 坐标 : ",datax,datax)
                  else:
                      a1=a1+1
      
          
      
          for c in img.find_circles(threshold = 3500, x_margin = 10, y_margin = 10, r_margin = 10,
                  r_min = 2, r_max = 100, 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 75<statistics.l_mode()<100 and -27<statistics.a_mode()<96 and -7<statistics.b_mode()<31:
                  img.draw_circle(c.x(), c.y(), c.r(), color = (255, 0, 0))
                  #识别到的白色圆形用红色的圆框出来
                  print("白",c.x(),c.y())
      
      
          for c in img.find_circles(threshold = 3500, x_margin = 10, y_margin = 10, r_margin = 10,
                  r_min = 2, r_max = 100, 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
              print(statistics)
              if 30<statistics.l_mode()<100 and 5<statistics.a_mode()<73 and -42<statistics.b_mode()<32:
                  img.draw_circle(c.x(), c.y(), c.r(), color = (255, 0, 0))
                  #识别到的粉色圆形用红色的圆框出来
                  print("粉",c.x(),c.y())
      


    • find_circles函数比较慢,你应该把103行,和115行,合并在一起。

      
          for c in img.find_circles(threshold = 3500, x_margin = 10, y_margin = 10, r_margin = 10,
                  r_min = 2, r_max = 100, 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 75<statistics.l_mode()<100 and -27<statistics.a_mode()<96 and -7<statistics.b_mode()<31:
                  img.draw_circle(c.x(), c.y(), c.r(), color = (255, 0, 0))
                  #识别到的白色圆形用红色的圆框出来
                  print("白",c.x(),c.y())
      
              if 30<statistics.l_mode()<100 and 5<statistics.a_mode()<73 and -42<statistics.b_mode()<32:
                  img.draw_circle(c.x(), c.y(), c.r(), color = (255, 0, 0))
                  #识别到的粉色圆形用红色的圆框出来
                  print("粉",c.x(),c.y())