导航

    • 登录
    • 搜索
    • 版块
    • 产品
    • 教程
    • 论坛
    • 淘宝
    1. 主页
    2. rgt4
    3. 楼层
    R
    • 举报资料
    • 资料
    • 关注
    • 粉丝
    • 屏蔽
    • 帖子
    • 楼层
    • 最佳
    • 群组

    rgt4 发布的帖子

    • 运行程序时报错咋处理

      从SD卡里导入文件进行特征点检测，提示运行内存不足，除了降低图片分辨率外，还有其他技巧可以解决这个问题吗？
      0_1590548509854_8cb4f6c44a18e6d14aad6c4ea263aca.png

      import sensor, time, image
from pid import PID
from pyb import Servo
pan_servo=Servo(1)
tilt_servo=Servo(2)
pan_pid = PID(p=0.10, i=0, imax=90)
tilt_pid = PID(p=0.10, i=0, imax=90)
sensor.reset()
sensor.set_contrast(3)
sensor.set_gainceiling(16)
sensor.set_framesize(sensor.VGA)
sensor.set_windowing((320, 240))
sensor.set_pixformat(sensor.GRAYSCALE)
sensor.skip_frames(time = 2000)
sensor.set_auto_gain(False, value=100)

def draw_keypoints(img, kpts):
    if kpts:
        print(kpts)
        img.draw_keypoints(kpts)
        img = sensor.snapshot()
        time.sleep(1000)
kpts1 = image.load_descriptor("/A1.orb")
kpts2 = image.load_descriptor("/A2.orb")
kpts3 = image.load_descriptor("/A3.orb")
kpts4 = image.load_descriptor("/A4.orb")
kpts5 = image.load_descriptor("/A5.orb")
kpts6 = image.load_descriptor("/A6.orb")
kpts7 = image.load_descriptor("/A7.orb")
kpts8 = image.load_descriptor("/A8.orb")
kpts9 = image.load_descriptor("/B1.orb")
kpts10 = image.load_descriptor("/B2.orb")
kpts11 = image.load_descriptor("/B3.orb")
kpts12 = image.load_descriptor("/B4.orb")
kpts13 = image.load_descriptor("/B5.orb")
kpts14 = image.load_descriptor("/B6.orb")
kpts15 = image.load_descriptor("/B7.orb")
kpts16 = image.load_descriptor("/B8.orb")
kpts17 = image.load_descriptor("/C1.orb")
kpts18 = image.load_descriptor("/C2.orb")
kpts19 = image.load_descriptor("/C3.orb")
kpts20 = image.load_descriptor("/C4.orb")
kpts21 = image.load_descriptor("/C5.orb")
kpts22 = image.load_descriptor("/C6.orb")
kpts23 = image.load_descriptor("/C7.orb")
kpts24 = image.load_descriptor("/C8.orb")
kpts25 = image.load_descriptor("/D1.orb")
kpts26 = image.load_descriptor("/D2.orb")
kpts27 = image.load_descriptor("/D3.orb")
kpts28 = image.load_descriptor("/D4.orb")
kpts29 = image.load_descriptor("/D5.orb")
kpts30 = image.load_descriptor("/D6.orb")
kpts31 = image.load_descriptor("/D7.orb")
kpts32 = image.load_descriptor("/D8.orb")
img = sensor.snapshot()
clock = time.clock()
while (True):
    clock.tick()
    img = sensor.snapshot()
    pan_servo.angle(0)
    time.sleep(200)
    tilt_servo.angle(-17)
    time.sleep(200)
    kptsA = img.find_keypoints(max_keypoints=150, threshold=10, normalized=True)
    if (kptsA):
        match1 = image.match_descriptor(kpts1, kptsA, threshold=85)
        match2 = image.match_descriptor(kpts2, kptsA, threshold=85)
        match3 = image.match_descriptor(kpts3, kptsA, threshold=85)
        match4 = image.match_descriptor(kpts4, kptsA, threshold=85)
        match5 = image.match_descriptor(kpts5, kptsA, threshold=85)
        match6 = image.match_descriptor(kpts6, kptsA, threshold=85)
        match7 = image.match_descriptor(kpts7, kptsA, threshold=85)
        match8 = image.match_descriptor(kpts8, kptsA, threshold=85)
        match9 = image.match_descriptor(kpts9, kptsA, threshold=85)
        match10 = image.match_descriptor(kpts10, kptsA, threshold=85)
        match11 = image.match_descriptor(kpts11, kptsA, threshold=85)
        match12 = image.match_descriptor(kpts12, kptsA, threshold=85)
        match13 = image.match_descriptor(kpts13, kptsA, threshold=85)
        match14 = image.match_descriptor(kpts14, kptsA, threshold=85)
        match15 = image.match_descriptor(kpts15, kptsA, threshold=85)
        match16 = image.match_descriptor(kpts16, kptsA, threshold=85)
        match17 = image.match_descriptor(kpts17, kptsA, threshold=85)
        match18 = image.match_descriptor(kpts18, kptsA, threshold=85)
        match19 = image.match_descriptor(kpts19, kptsA, threshold=85)
        match20 = image.match_descriptor(kpts20, kptsA, threshold=85)
        match21 = image.match_descriptor(kpts21, kptsA, threshold=85)
        match22 = image.match_descriptor(kpts22, kptsA, threshold=85)
        match23 = image.match_descriptor(kpts23, kptsA, threshold=85)
        match24 = image.match_descriptor(kpts24, kptsA, threshold=85)
        match25 = image.match_descriptor(kpts25, kptsA, threshold=85)
        match26 = image.match_descriptor(kpts26, kptsA, threshold=85)
        match27 = image.match_descriptor(kpts27, kptsA, threshold=85)
        match28 = image.match_descriptor(kpts28, kptsA, threshold=85)
        match29 = image.match_descriptor(kpts29, kptsA, threshold=85)
        match30 = image.match_descriptor(kpts30, kptsA, threshold=85)
        match31 = image.match_descriptor(kpts31, kptsA, threshold=85)
        match32 = image.match_descriptor(kpts32, kptsA, threshold=85)
        if (match1.count()>10):
            pan_servo.angle(40)
            time.sleep(400)
            pan_servo.angle(0)
            time.sleep(400)
        elif (match2.count()>10):
            pan_servo.angle(40)
            time.sleep(400)
            pan_servo.angle(0)
            time.sleep(400)
        elif (match3.count()>10):
            pan_servo.angle(40)
            time.sleep(400)
            pan_servo.angle(0)
            time.sleep(400)
        elif (match4.count()>10):
            pan_servo.angle(40)
            time.sleep(400)
            pan_servo.angle(0)
            time.sleep(400)
        elif (match5.count()>10):
            pan_servo.angle(40)
            time.sleep(400)
            pan_servo.angle(0)
            time.sleep(400)
        elif (match6.count()>10):
            pan_servo.angle(40)
            time.sleep(400)
            pan_servo.angle(0)
            time.sleep(400)
        elif (match7.count()>10):
            pan_servo.angle(40)
            time.sleep(400)
            pan_servo.angle(0)
            time.sleep(400)
        elif (match8.count()>10):        
            pan_servo.angle(40)
            time.sleep(400)
            pan_servo.angle(0)
            time.sleep(400)            
        elif (match9.count()>10):
            tilt_servo.angle(-47)
            time.sleep(400)
            tilt_servo.angle(-17)
            time.sleep(400)
        elif (match10.count()>10):
            tilt_servo.angle(-47)
            time.sleep(400)
            tilt_servo.angle(-17)
            time.sleep(400)
        elif (match11.count()>10):
            tilt_servo.angle(-47)
            time.sleep(400)
            tilt_servo.angle(-17)
            time.sleep(400)
        elif (match12.count()>10):
            tilt_servo.angle(-47)
            time.sleep(400)
            tilt_servo.angle(-17)
            time.sleep(400)
        elif (match13.count()>10):
            tilt_servo.angle(-47)
            time.sleep(400)
            tilt_servo.angle(-17)
            time.sleep(400)
        elif (match14.count()>10):
            tilt_servo.angle(-47)
            time.sleep(400)
            tilt_servo.angle(-17)
            time.sleep(400)
        elif (match15.count()>10):
            tilt_servo.angle(-47)
            time.sleep(400)
            tilt_servo.angle(-17)
            time.sleep(400)
        elif (match16.count()>10):
            tilt_servo.angle(-47)
            time.sleep(400)
            tilt_servo.angle(-17)
            time.sleep(400)
        elif (match17.count()>10):
            pan_servo.angle(-35)
            time.sleep(400)
            pan_servo.angle(0)
            time.sleep(400)
        elif (match18.count()>10):
            pan_servo.angle(-35)
            time.sleep(400)
            pan_servo.angle(0)
            time.sleep(400)
        elif (match19.count()>10):
            pan_servo.angle(-35)
            time.sleep(400)
            pan_servo.angle(0)
            time.sleep(400)
        elif (match20.count()>10):
            pan_servo.angle(-35)
            time.sleep(400)
            pan_servo.angle(0)
            time.sleep(400)
        elif (match21.count()>10):
            pan_servo.angle(-35)
            time.sleep(400)
            pan_servo.angle(0)
            time.sleep(400)
        elif (match22.count()>10):        
            pan_servo.angle(-35)
            time.sleep(400)
            pan_servo.angle(0)
            time.sleep(400)
        elif (match23.count()>10):        
            pan_servo.angle(-35)
            time.sleep(400)
            pan_servo.angle(0)
            time.sleep(400)
        elif (match24.count()>10):
            pan_servo.angle(-35)
            time.sleep(400)
            pan_servo.angle(0)
            time.sleep(400)            
        elif (match25.count()>10):
            tilt_servo.angle(13)
            time.sleep(400)
            tilt_servo.angle(-17)
            time.sleep(400)
        elif (match26.count()>10):
            tilt_servo.angle(13)
            time.sleep(400)
            tilt_servo.angle(-17)
            time.sleep(400)
        elif (match27.count()>10):
            tilt_servo.angle(13)
            time.sleep(400)
            tilt_servo.angle(-17)
            time.sleep(400)
        elif (match28.count()>10):
            tilt_servo.angle(13)
            time.sleep(400)
            tilt_servo.angle(-17)
            time.sleep(400)
        elif (match29.count()>10):
            tilt_servo.angle(13)
            time.sleep(400)
            tilt_servo.angle(-17)
            time.sleep(400)
        elif (match30.count()>10):
            tilt_servo.angle(13)
            time.sleep(400)
            tilt_servo.angle(-17)
            time.sleep(400)
        elif (match31.count()>10):        
            tilt_servo.angle(13)
            time.sleep(400)
            tilt_servo.angle(-17)
            time.sleep(400)
        elif (match32.count()>10):
            tilt_servo.angle(13)
            time.sleep(400)
            tilt_servo.angle(-17)
            time.sleep(400)    
        else :
            pan_servo.angle(0)
            time.sleep(200)
            tilt_servo.angle(-17)
            time.sleep(200)
        print(kptsA, "matched:%d dt:%d"%(match4.count(), match4.theta()))
    img.draw_string(0, 0, "FPS:%.2f"%(clock.fps()))
      发布在 OpenMV Cam
      R
      rgt4
    • 找不到DfuseDemo.exe在哪里下载

      下载IDE的时候，只有IDE.exe，并没有OpenMV IDE的一个大文件夹，因此无法如教程里那样找到烧录软件和最新固件，请给一个下载链接。0_1589425519596_777.PNG 0_1589425534504_888.PNG

      发布在 OpenMV Cam
      R
      rgt4
    • RE: OpenMV变砖了，无法DFU升级？

      @tgne 您好，最后怎么解决的？

      发布在 OpenMV Cam
      R
      rgt4
    • 接上一个摄像头失效的问题

      为什么open mv连着电脑时，如果电脑没电关机了，充电后再次连接open mv 时就连接不了了？是电脑关机时有电信号传给摄像头，然后把它烧坏了吗？我测摄像头stm32的电压时，依旧还有3.3V。

      发布在 OpenMV Cam
      R
      rgt4
    • RE: USB识别不了，电脑没电后再开。不行，

      你好，最后咋解决的？

      发布在 OpenMV Cam
      R
      rgt4
    • RE: USB识别不了，电脑没电后再开。不行，

      @ovmz 你好，这个问题解决了吗？咋解决的？

      发布在 OpenMV Cam
      R
      rgt4
    • Openmv Cam用着用着突然就坏了

      连着openmv cam的时候，电脑没电关机了，我记得是点了重置之后电脑才关机的，但是没拔数据线。然后电脑充电后再次连接Cam的时候，就识别不了openmv cam了。我按照你在论坛里发的那个检查驱动、检查数据线、升级固件、dfu啥的操作了，感觉试了没用，而且重点是，由于由于无法识别openmv cam，感觉它的重修编码并没有被执行进去，连接跳线也感觉并没有进行编码，只是摄像头发热的厉害。
      还有什么方法吗？如果寄回维修，咋收费？去年九月份买的。0_1589380932462_IMG20200513200804.jpg

      发布在 OpenMV Cam
      R
      rgt4
    • 边缘检测的不兼容性问题

      想对经过【边缘检测】处理后得到的图片进行【特征点保存】，但是运行代码时，【边缘检测】只能进行一次，无法循环，而【特征点保存】可以一直循环检测下去。想让【边缘检测】和【特征点保存】一起循环运行，

      import sensor, time, image

sensor.reset()

sensor.set_contrast(3)
sensor.set_gainceiling(16)
sensor.set_framesize(sensor.VGA)
sensor.set_windowing((320, 240))
sensor.set_pixformat(sensor.GRAYSCALE)

sensor.skip_frames(time = 2000)
sensor.set_auto_gain(False, value=100)

FILE_NAME = "666"
img = sensor.snapshot()
kernel_size = 1 # kernel width = (size*2)+1, kernel height = (size*2)+1
kernel = [-1, -1, -1,\
          -1, +8, -1,\
          -1, -1, -1]
thresholds = [(100, 255)] # grayscale thresholds设置阈值

clock = time.clock() # 追踪FPS

if (sensor.get_id() == sensor.OV7725):
    sensor.__write_reg(0xAC, 0xDF)
    sensor.__write_reg(0x8F, 0xFF)

while(True):
    clock.tick() # Track elapsed milliseconds between snapshots().
    img = sensor.snapshot() # Take a picture and return the image.

    img.morph(kernel_size, kernel)
    img.binary(thresholds)
    img.erode(1, threshold = 2)
    print(clock.fps()) # Note: Your OpenMV Cam runs about half as fast while
    time.sleep(100)
    
    kpts = img.find_keypoints(max_keypoints=150, threshold=10, scale_factor=1.2)
    if (kpts == None):
        raise(Exception("Couldn't find any keypoints!"))
    
    image.save_descriptor(kpts, "/%s.orb"%(FILE_NAME))
    
    img.draw_keypoints(kpts)
    sensor.snapshot()

      代码该怎么改？谢谢

      发布在 OpenMV Cam
      R
      rgt4
    • RE: 为啥用保存的差距这么大？

      @kidswong999 换了白底，但还是一样，摆放角度一变就基本识别不出来。还有其他办法吗？

      发布在 OpenMV Cam
      R
      rgt4
    • 为啥用保存的差距这么大？

      在进行特征点检测时，物体A的特征是从文件导入的，然后用它来和实时检测到的物体进行相似度对比。
      我发现，当我用image.save_descriptor保存好物体A的特征后，不去移动物体A，就立马用物体A和刚才它自身保存进去的特征文件进行对比，这时匹配度很高，但是只要稍微移动一下物体A，它的匹配度就极低。根本没有任何的光线环境改变，而且特征点检测是可以改变样品摆放角度的吧，我记得去年调试的时候，只要把样品特征保存进SD卡，然后无论怎么改变角度，匹配度都很高，我也没改代码，但现在感觉特征点检测用起来跟模板匹配一样死板了，不知道该怎么解决。

      发布在 OpenMV Cam
      R
      rgt4
    • 为啥起始定义的是低电平，但连接IDE时，未运行代码时输出的是高电平？

      起始定义的是低电平，如下：
      p_out = Pin('P3', Pin.OUT_PP)
      p_out.low()
      在运行代码时，可以按照指令输出高电平或低电平，但是当我的电脑和Openmv Cam相连时，也就是仅仅进行连接，但没有点击IDE左下角的三角形运行时，Openmv Cam持续输出高电平。

      import sensor, time, image
from pid import PID
from pyb import Servo
from pyb import Pin
pan_servo=Servo(1)
tilt_servo=Servo(2)
pan_pid = PID(p=0.15, i=0, imax=90)
tilt_pid = PID(p=0.15, i=0, imax=90)
sensor.reset()
sensor.set_contrast(3)
sensor.set_gainceiling(16)
sensor.set_framesize(sensor.VGA)
sensor.set_windowing((320, 240))
sensor.set_pixformat(sensor.GRAYSCALE)
sensor.skip_frames(time = 2000)
sensor.set_auto_gain(False, value=100)
p_out = Pin('P3', Pin.OUT_PP)
p_out.low()

def draw_keypoints(img, kpts):
    if kpts:
        print(kpts)
        img.draw_keypoints(kpts)
        img = sensor.snapshot()
        time.sleep(1000)
kpts1 = image.load_descriptor("/Unrecyclable.orb")
img = sensor.snapshot()
clock = time.clock()
while (True):
    clock.tick()
    img = sensor.snapshot()
    kptsA = img.find_keypoints(max_keypoints=150, threshold=10, normalized=True)
    if (kptsA):
        match1 = image.match_descriptor(kpts1, kptsA, threshold=85)
        if (match1.count()>10):
            p_out.high()
        else :
            p_out.low()           
        print(kptsA, "matched:%d dt:%d"%(match1.count(), match1.theta()))
    img.draw_string(0, 0, "FPS:%.2f"%(clock.fps()))
      发布在 OpenMV Cam
      R
      rgt4
    • 关于模板匹配之外的算法

      除了模板匹配这类比较死板的方式之外，有哪些方法可以对比两张图片的相似度？

      发布在 OpenMV Cam
      R
      rgt4
    • 预处理的结果如何加以保存

      边缘检测相当于预处理，得到的图像如何保存到openmv里。

      发布在 OpenMV Cam
      R
      rgt4
    • 请问如何实现这个功能：用特征点检测例程检测到有物体存在时，让openmv输出一个高电平，这个代码怎么写？

      请问如何实现这个功能：用特征点检测例程检测到有物体存在时，让openmv输出一个高电平，这个代码怎么写？

      发布在 OpenMV Cam
      R
      rgt4