In this video lesson we show you how to install DLIB and face_recognition libraries on the NVIDIA Jetson Xavier NX. We take you through the installs step-by-step. This will be foundational libraries for future lessons on Face Recognition and Deep Learning.
Robotics Training LESSON 9: Calibrate Robot Car Turn Angles with Linear Regression
Guys in this lesson we show how you can use linear regression to more accurately calibrate turns for our robot car. We will take you through the math step-by-step, and then show how to code up the results. If you want to play along at home, you can get your Elegoo Robot Car HERE.
For your convenience, the code we developed in the lesson is included below.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 | int ENA=5; int ENB=6; int IN1=7; int IN2=8; int IN3=9; int IN4=11; float d; int degRot; int left; int right; float v; void setup() { // put your setup code here, to run once: Serial.begin(9600); pinMode(ENA,OUTPUT); pinMode(ENB,OUTPUT); pinMode(IN1,OUTPUT); pinMode(IN2,OUTPUT); pinMode(IN3,OUTPUT); pinMode(IN4,OUTPUT); digitalWrite(ENA,HIGH); digitalWrite(ENB,HIGH); } void loop() { int wv; v=1.2; wv=(v-.35)/.0075; left=wv; right=wv; setSpeed(left,right); forward(5, v); turnRight(180, wv); forward(5, v); //calR(wv); //forward(8,v); //v=1.5; //wv=(v-.35)/.0075; //left=wv; //right=wv; //setSpeed(left,right); //backward(8,v); while(1==1){ } } void setSpeed(int leftVal,int rightVal){ analogWrite(ENA,leftVal); analogWrite(ENB,rightVal); } void forward(float d, float v){ float t; digitalWrite(IN1,HIGH); digitalWrite(IN2,LOW); digitalWrite(IN3,LOW); digitalWrite(IN4,HIGH); t=d/v*1000; delay(t); stopCar(); } void backward(float d, float v){ float t; digitalWrite(IN1,LOW); digitalWrite(IN2,HIGH); digitalWrite(IN3,HIGH); digitalWrite(IN4,LOW); t=d/v*1000; delay(t); stopCar(); } void turnRight(int deg, int wv){ float t; stopCar(); delay(100); analogWrite(ENA,125); analogWrite(ENB,125); digitalWrite(IN1,HIGH); digitalWrite(IN2,LOW); digitalWrite(IN3,HIGH); digitalWrite(IN4,LOW); t=(deg+6)/136.29*1000.; Serial.println(deg); delay(t); stopCar(); analogWrite(ENA,wv); analogWrite(ENB,wv); } void turnLeft(int deg, int wv){ float t; analogWrite(ENA,125); analogWrite(ENB,125); digitalWrite(IN1,LOW); digitalWrite(IN2,HIGH); digitalWrite(IN3,LOW); digitalWrite(IN4,HIGH); t=(deg+6)/136.29*1000.; Serial.println(deg); Serial.println(deg); delay(t); stopCar(); analogWrite(ENA,wv); analogWrite(ENB,wv); } void stopCar(){ digitalWrite(IN1,LOW); digitalWrite(IN2,LOW); digitalWrite(IN3,LOW); digitalWrite(IN4,LOW); } void calF(){ digitalWrite(IN1,HIGH); digitalWrite(IN2,LOW); digitalWrite(IN3,LOW); digitalWrite(IN4,HIGH); delay(5000); stopCar(); } void calB(){ digitalWrite(IN1,LOW); digitalWrite(IN2,HIGH); digitalWrite(IN3,HIGH); digitalWrite(IN4,LOW); delay(5000); stopCar(); } void calR(int wv){ stopCar(); analogWrite(ENA,125); analogWrite(ENB,125); digitalWrite(IN1,HIGH); digitalWrite(IN2,LOW); digitalWrite(IN3,HIGH); digitalWrite(IN4,LOW); delay(3000); analogWrite(ENA,wv); analogWrite(ENB,wv); stopCar(); } void calL(int wv){ analogWrite(ENA,125); analogWrite(ENB,125); digitalWrite(IN1,LOW); digitalWrite(IN2,HIGH); digitalWrite(IN3,LOW); digitalWrite(IN4,HIGH); delay(5000); analogWrite(ENA,wv); analogWrite(ENB,wv); stopCar(); } |
Jetson Xavier NX Lesson 12: Intelligent Scanning for Objects of Interest
In this Video Tutorial we show how a camera on a pan/tilt control system can be programmed to search for an object of interest, and then track it when found. Our system has two independent camera systems, and each can track a separate item of interest independently. The code is written in python, using the OpenCV library. The video takes you through the lesson step-by-step, and then the code is included below for your convenience.
If you want to play along at home, we are using the Jetson Xavier NX, which you can pick up HERE. You will also need to of the bracket/servo kits, which you can get HERE, and then two Raspberry Pi Version two cameras, available HERE.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 | import cv2 import numpy as np import time from adafruit_servokit import ServoKit print(cv2.__version__) timeMark=time.time() dtFIL=0 scanRight=True scanLeft=True def nothing(x): pass cv2.namedWindow('TrackBars') cv2.moveWindow('TrackBars',1320,0) cv2.createTrackbar('hueLower', 'TrackBars',100,179,nothing) cv2.createTrackbar('hueUpper', 'TrackBars',116,179,nothing) cv2.createTrackbar('satLow', 'TrackBars',160,255,nothing) cv2.createTrackbar('satHigh', 'TrackBars',255,255,nothing) cv2.createTrackbar('valLow', 'TrackBars',150,255,nothing) cv2.createTrackbar('valHigh', 'TrackBars',255,255,nothing) cv2.namedWindow('TrackBars2') cv2.moveWindow('TrackBars2',1100,0) cv2.createTrackbar('hueLower2', 'TrackBars2',150,179,nothing) cv2.createTrackbar('hueUpper2', 'TrackBars2',170,179,nothing) cv2.createTrackbar('satLow2', 'TrackBars2',160,255,nothing) cv2.createTrackbar('satHigh2', 'TrackBars2',255,255,nothing) cv2.createTrackbar('valLow2', 'TrackBars2',150,255,nothing) cv2.createTrackbar('valHigh2', 'TrackBars2',255,255,nothing) kit=ServoKit(channels=16) tilt1=90 pan1=90 tilt2=90 pan2=90 dPan1=1 dPan2=1 dTilt1=10 dTilt2=10 kit.servo[0].angle=pan1 kit.servo[1].angle=tilt1 kit.servo[2].angle=pan2 kit.servo[3].angle=tilt2 width=720 height=480 flip=2 font=cv2.FONT_HERSHEY_SIMPLEX camSet1='nvarguscamerasrc sensor-id=0 ee-mode=1 ee-strength=0 tnr-mode=2 tnr-strength=1 wbmode=3 ! video/x-raw(memory:NVMM), width=3264, height=2464, framerate=21/1,format=NV12 ! nvvidconv flip-method='+str(flip)+' ! video/x-raw, width='+str(width)+', height='+str(height)+', format=BGRx ! videoconvert ! video/x-raw, format=BGR ! videobalance contrast=1.3 brightness=-.2 saturation=1.2 ! appsink drop=True' camSet2='nvarguscamerasrc sensor-id=1 ee-mode=1 ee-strength=0 tnr-mode=2 tnr-strength=1 wbmode=3 ! video/x-raw(memory:NVMM), width=3264, height=2464, framerate=21/1,format=NV12 ! nvvidconv flip-method='+str(flip)+' ! video/x-raw, width='+str(width)+', height='+str(height)+', format=BGRx ! videoconvert ! video/x-raw, format=BGR ! videobalance contrast=1.3 brightness=-.2 saturation=1.2 ! appsink drop=True' #camSet='nvarguscamerasrc sensor-id=0 ! video/x-raw(memory:NVMM), width=3264, height=2464, framerate=21/1,format=NV12 ! nvvidconv flip-method='+str(flip)+' ! video/x-raw, width='+str(width)+', height='+str(height)+', format=BGRx ! videoconvert ! video/x-raw, format=BGR ! appsink' #camSet ='v4l2src device=/dev/video1 ! video/x-raw,width='+str(width)+',height='+str(height)+',framerate=20/1 ! videoconvert ! appsink' cam1=cv2.VideoCapture(camSet1) cam2=cv2.VideoCapture(camSet2) while True: _, frame1 = cam1.read() _, frame2 = cam2.read() hsv1=cv2.cvtColor(frame1,cv2.COLOR_BGR2HSV) hsv2=cv2.cvtColor(frame2,cv2.COLOR_BGR2HSV) hueLow=cv2.getTrackbarPos('hueLower', 'TrackBars') hueUp=cv2.getTrackbarPos('hueUpper', 'TrackBars') Ls=cv2.getTrackbarPos('satLow', 'TrackBars') Us=cv2.getTrackbarPos('satHigh', 'TrackBars') Lv=cv2.getTrackbarPos('valLow', 'TrackBars') Uv=cv2.getTrackbarPos('valHigh', 'TrackBars') l_b=np.array([hueLow,Ls,Lv]) u_b=np.array([hueUp,Us,Uv]) hueLow2=cv2.getTrackbarPos('hueLower2', 'TrackBars2') hueUp2=cv2.getTrackbarPos('hueUpper2', 'TrackBars2') Ls2=cv2.getTrackbarPos('satLow2', 'TrackBars2') Us2=cv2.getTrackbarPos('satHigh2', 'TrackBars2') Lv2=cv2.getTrackbarPos('valLow2', 'TrackBars2') Uv2=cv2.getTrackbarPos('valHigh2', 'TrackBars2') l_b2=np.array([hueLow2,Ls2,Lv2]) u_b2=np.array([hueUp2,Us2,Uv2]) FGmask1=cv2.inRange(hsv1,l_b,u_b) FGmask2=cv2.inRange(hsv2,l_b2,u_b2) cv2.imshow('FGmask1',FGmask1) cv2.moveWindow('FGmask1',0,0) cv2.imshow('FGmask2',FGmask2) cv2.moveWindow('FGmask2',350,0) contours1,_ = cv2.findContours(FGmask1,cv2.RETR_EXTERNAL,cv2.CHAIN_APPROX_SIMPLE) contours1=sorted(contours1,key=lambda x:cv2.contourArea(x),reverse=True) for cnt in contours1: area=cv2.contourArea(cnt) (x,y,w,h)=cv2.boundingRect(cnt) if area>=100: scanLeft=False cv2.rectangle(frame1,(x,y),(x+w,y+h),(0,255,255),3) objX=x+w/2 objY=y+h/2 errorPan1=objX-width/2 errorTilt1=objY-height/2 if abs(errorPan1)>15: pan1=pan1+errorPan1/40 if abs(errorTilt1)>15: tilt1=tilt1-errorTilt1/40 if pan1>180: pan1=180 print('Pan Out of Range') if pan1<0: pan1=0 print('Pan Out of Range') if tilt1>180: tilt1=180 print('Tilt Out of Range') if tilt1<0: tilt1=0 kit.servo[2].angle=pan1 kit.servo[3].angle=tilt1 break contours2,_ = cv2.findContours(FGmask2,cv2.RETR_EXTERNAL,cv2.CHAIN_APPROX_SIMPLE) contours2=sorted(contours2,key=lambda x:cv2.contourArea(x),reverse=True) for cnt in contours2: area=cv2.contourArea(cnt) (x,y,w,h)=cv2.boundingRect(cnt) if area>=100: scanRight=False cv2.rectangle(frame2,(x,y),(x+w,y+h),(0,255,255),3) objX=x+w/2 objY=y+h/2 errorPan2=objX-width/2 errorTilt2=objY-height/2 if abs(errorPan2)>15: pan2=pan2+errorPan2/40 if abs(errorTilt2)>15: tilt2=tilt2-errorTilt2/40 if pan2>180: pan2=180 print('Pan Out of Range') if pan2<0: pan2=0 print('Pan Out of Range') if tilt2>180: tilt2=180 print('Tilt Out of Range') if tilt2<0: tilt2=0 kit.servo[0].angle=pan2 kit.servo[1].angle=tilt2 break if scanLeft==True: if pan1>=179: dPan1=abs(dPan1)*(-1) if pan1<=1: dPan1=abs(dPan1) if pan1>=179 or pan1<=1: if tilt1>=170: dTilt1=abs(dTilt1)*(-1) if tilt1<=10: dTilt1=abs(dTilt1) tilt1=tilt1+dTilt1 pan1=pan1+dPan1 kit.servo[2].angle=pan1 kit.servo[3].angle=tilt1 scanLeft=True if scanRight==True: if pan2>=179: dPan2=abs(dPan2)*(-1) if pan2<=1: dPan2=abs(dPan2) if pan2>=179 or pan2<=1: if tilt2>=170: dTilt2=abs(dTilt2)*(-1) if tilt2<=10: dTilt2=abs(dTilt2) tilt2=tilt2+dTilt2 pan2=pan2+dPan2 kit.servo[0].angle=pan2 kit.servo[1].angle=tilt2 scanRight=True frame3=np.hstack((frame1,frame2)) dt=time.time()-timeMark timeMark=time.time() dtFIL=.9*dtFIL + .1*dt fps=1/dtFIL cv2.rectangle(frame3,(0,0),(150,40),(0,0,255),-1) cv2.putText(frame3,'fps: '+str(round(fps,1)),(0,30),font,1,(0,255,255),2) #cv2.imshow('myCam1',frame1) #cv2.imshow('myCam2',frame2) cv2.imshow('comboCam',frame3) cv2.moveWindow('comboCam',0,450) if cv2.waitKey(1)==ord('q'): break cam1.release() cam2.release() cv2.destroyAllWindows() |
AI on the Jetson Nano LESSON 53: Object Detection and Recognition in OpenCV
In this video lesson we learn how to use the NVIDIA Jetson Inference tools for detect objects in a live video. The software developed in this lesson is included below for your convenience.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 | import jetson.inference import jetson.utils import time import cv2 import numpy as np timeStamp=time.time() fpsFilt=0 net=jetson.inference.detectNet('ssd-mobilenet-v2',threshold=.5) dispW=1280 dispH=720 flip=2 font=cv2.FONT_HERSHEY_SIMPLEX # Gstreamer code for improvded Raspberry Pi Camera Quality #camSet='nvarguscamerasrc wbmode=3 tnr-mode=2 tnr-strength=1 ee-mode=2 ee-strength=1 ! video/x-raw(memory:NVMM), width=3264, height=2464, format=NV12, framerate=21/1 ! nvvidconv flip-method='+str(flip)+' ! video/x-raw, width='+str(dispW)+', height='+str(dispH)+', format=BGRx ! videoconvert ! video/x-raw, format=BGR ! videobalance contrast=1.5 brightness=-.2 saturation=1.2 ! appsink' #cam=cv2.VideoCapture(camSet) #cam=jetson.utils.gstCamera(dispW,dispH,'0') cam=cv2.VideoCapture('/dev/video1') cam.set(cv2.CAP_PROP_FRAME_WIDTH, dispW) cam.set(cv2.CAP_PROP_FRAME_HEIGHT, dispH) #cam=jetson.utils.gstCamera(dispW,dispH,'/dev/video1') #display=jetson.utils.glDisplay() #while display.IsOpen(): while True: #img, width, height= cam.CaptureRGBA() _,img = cam.read() height=img.shape[0] width=img.shape[1] frame=cv2.cvtColor(img,cv2.COLOR_BGR2RGBA).astype(np.float32) frame=jetson.utils.cudaFromNumpy(frame) detections=net.Detect(frame, width, height) for detect in detections: #print(detect) ID=detect.ClassID top=detect.Top left=detect.Left bottom=detect.Bottom right=detect.Right item=net.GetClassDesc(ID) print(item,top,left,bottom,right) #display.RenderOnce(img,width,height) dt=time.time()-timeStamp timeStamp=time.time() fps=1/dt fpsFilt=.9*fpsFilt + .1*fps #print(str(round(fps,1))+' fps') cv2.putText(img,str(round(fpsFilt,1))+' fps',(0,30),font,1,(0,0,255),2) cv2.imshow('detCam',img) cv2.moveWindow('detCam',0,0) if cv2.waitKey(1)==ord('q'): break cam.release() cv2.destroyAllWindows() |
Robotics Training LESSON 8: Setting Speed of the Smart Car
In this lesson we show how to program the speed of the Elegoo Smart Car Version 3.0 using the Infrared (IR) remote control. We take you through the process of using the remote step-by-step, and for your convenience we include the code below. If you want to play along at home, you can pick up your gear HERE.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 | int ENA=5; int ENB=6; int IN1=7; int IN2=8; int IN3=9; int IN4=11; float d; int degRot; int left; int right; int wv; float v; void setup() { // put your setup code here, to run once: Serial.begin(9600); pinMode(ENA,OUTPUT); pinMode(ENB,OUTPUT); pinMode(IN1,OUTPUT); pinMode(IN2,OUTPUT); pinMode(IN3,OUTPUT); pinMode(IN4,OUTPUT); digitalWrite(ENA,HIGH); digitalWrite(ENB,HIGH); } void loop() { v=1.5; wv=(v-.35)/.0075; left=wv; right=wv; setSpeed(left,right); forward(8,v); v=1.5; wv=(v-.35)/.0075; left=wv; right=wv; setSpeed(left,right); backward(8,v); while(1==1){ } } void setSpeed(int leftVal,int rightVal){ analogWrite(ENA,leftVal); analogWrite(ENB,rightVal); } void forward(float d, float v){ float t; digitalWrite(IN1,HIGH); digitalWrite(IN2,LOW); digitalWrite(IN3,LOW); digitalWrite(IN4,HIGH); t=d/v*1000; delay(t); stopCar(); } void backward(float d, float v){ float t; digitalWrite(IN1,LOW); digitalWrite(IN2,HIGH); digitalWrite(IN3,HIGH); digitalWrite(IN4,LOW); t=d/v*1000; delay(t); stopCar(); } void turnRight(int deg){ float t; digitalWrite(IN1,HIGH); digitalWrite(IN2,LOW); digitalWrite(IN3,HIGH); digitalWrite(IN4,LOW); t=deg/345.*1000.; Serial.println(deg); delay(t); stopCar(); } void turnLeft(float deg){ float t; digitalWrite(IN1,LOW); digitalWrite(IN2,HIGH); digitalWrite(IN3,LOW); digitalWrite(IN4,HIGH); t=deg/345.*1000.; Serial.println(deg); delay(t); stopCar(); } void stopCar(){ digitalWrite(IN1,LOW); digitalWrite(IN2,LOW); digitalWrite(IN3,LOW); digitalWrite(IN4,LOW); } void calF(){ digitalWrite(IN1,HIGH); digitalWrite(IN2,LOW); digitalWrite(IN3,LOW); digitalWrite(IN4,HIGH); delay(5000); stopCar(); } void calB(){ digitalWrite(IN1,LOW); digitalWrite(IN2,HIGH); digitalWrite(IN3,HIGH); digitalWrite(IN4,LOW); delay(5000); stopCar(); } void calR(){ digitalWrite(IN1,HIGH); digitalWrite(IN2,LOW); digitalWrite(IN3,HIGH); digitalWrite(IN4,LOW); delay(5000); stopCar(); } void calL(){ digitalWrite(IN1,LOW); digitalWrite(IN2,HIGH); digitalWrite(IN3,LOW); digitalWrite(IN4,HIGH); delay(5000); stopCar(); } |