矩形识别后，再识别颜色异常

[2mdu]

为啥识别到矩形了，但是没有进颜色值的范围判断。是什么原因呀
代码如下：
row_data=[-1,-1,-1,-1]
global area
for r in img.find_rects(threshold=10000):
area=(r[0], r[1], r[2], r[3])
statistics = img.get_statistics(roi=area)
xq= r[2] / r[3]
if 0<statistics.l_mode()<100 and 0<statistics.a_mode()<127 and 0<statistics.b_mode()<127:
row_data[0]= 1
img.draw_rectangle(r.rect(),color = (255, 255, 255))
elif 0<statistics.l_mode()<100 and -128<statistics.a_mode()<-23 and -128<statistics.b_mode()<127:
row_data[0]= 3
img.draw_rectangle(r.rect(),color = (255, 255, 255))
elif 0<statistics.l_mode()<100 and -128<statistics.a_mode()<127 and -128<statistics.b_mode()<-2:
row_data[0]= 2
img.draw_rectangle(r.rect(),color = (255, 255, 255))
row_data[1]=1

[kidswong999]

如果涉及代码，需要报错提示与全部代码文本，请注意不要贴代码图片

[2mdu]

statistics 怎么获取范围值，比如图中 的红色和蓝色啊

[2mdu]

import sensor, image, time
from pyb import UART


sensor.reset()
sensor.set_pixformat(sensor.RGB565)
sensor.set_framesize(sensor.QVGA)
sensor.skip_frames(time = 1000)
sensor.set_auto_gain(False)
sensor.set_auto_whitebal(False)
clock = time.clock()



thresholds = [(16, 71, 15, 102, -23, 97), # red_thresholds
              (11, 29, -33, -17, 0, 48), # green_thresholds
              (0, 42, -23, 16, -46, -5)] # blue_thresholds

shape = 1
threshold_index =0
area = 0
uart = UART(3, 115200)


def sending_data(cx,cy):
    global uart;
    data = ustruct.pack("<bbhhb",
                   0x2C,
                   0x12,
                   int(cx),
                   int(cy),
                   0x5B)
    uart.write(data);


def receive_data():
    global uart
    if uart.any():
        tmp_data = uart.readline();
        print("====接收===",tmp_data)
        return tmp_data;



def detect_shape(shape):
       row_data=[-1,-1]
       global area
  #  if shape==1:
       for r in img.find_rects(threshold=10000):
           area=(r[0], r[1], r[2], r[3])
           statistics = img.get_statistics(roi=area)
           xq= r[2] / r[3]
           if 0<statistics.l_mode()<100 and 20<statistics.a_mode()<127 and 10<statistics.b_mode()<127 and 0.8<xq<1.2:
              row_data[0]= 1
           elif 0<statistics.l_mode()<100 and -128<statistics.a_mode()<-23 and -128<statistics.b_mode()<127 and 0.8<xq<1.2:
             row_data[0]= 3
           elif 0<statistics.l_mode()<100 and -128<statistics.a_mode()<127 and -128<statistics.b_mode()<-2 and 0.8<xq<1.2:
              row_data[0]= 2
           row_data[1]=1
           img.draw_rectangle(r.rect(), color = (255, 255, 255))
  #  elif shape==2:
       for c in img.find_circles(threshold =2000, 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)
           if 0<statistics.l_mode()<100 and -128<statistics.a_mode()<0 and 0<statistics.b_mode()<127:#红色
               row_data[0]= 1
           elif 0<statistics.l_mode()<100 and -128<statistics.a_mode()<0 and 0<statistics.b_mode()<127:#绿色
               row_data[0]= 3
           elif 0<statistics.l_mode()<100 and -128<statistics.a_mode()<127 and -128<statistics.b_mode()<-2:#蓝色
                row_data[0]= 2
           row_data[1]=2
           img.draw_circle(c.x(), c.y(), c.r(), color = (255, 255, 255))
   # elif shape==3:
      #    for l in img.find_line_segments(roi=[60,40,150,150],merge_distance = 10, max_theta_diff = 10):
       #      img.draw_line(l.line(), color = (0, 0, 255))
       return row_data



while(True):
    clock.tick()
   # data= receive_data();
    img = sensor.snapshot()
    result =detect_shape(shape)
    print(clock.fps())

[2mdu]

识别视频中的形状率极低，这样情况怎么处理啊

[kidswong999]

你的代码我运行没有报错，你的问题是什么?

[2mdu]

@kidswong999 识别不出来矩形

[2mdu]

[kidswong999]

那看这个教程：https://book.openmv.cc/example/09-Feature-Detection/find-rects.html

[2mdu]

我看代码没有差别，找不到什么原因导致得。