1 00:00:02,000 --> 00:00:03,000 Hello everyone. 2 00:00:03,000 --> 00:00:09,000 In this video tutorial we will see how we can plot the detected objects after doing detections using 3 00:00:09,000 --> 00:00:14,000 YOLO and tracking using your deep sort of algorithm. 4 00:00:14,000 --> 00:00:21,000 So after doing detections using YOLO V8, we will blurred the detected objects and for object tracking 5 00:00:21,000 --> 00:00:23,000 we will use the deepsort algorithm. 6 00:00:23,000 --> 00:00:29,000 So for this project I will be using this GitHub repo over here so you can see this YOLO v8 Deepsort 7 00:00:29,000 --> 00:00:31,000 object tracking repo. 8 00:00:31,000 --> 00:00:40,000 So in this code we have already in this GitHub repo, the YOLO V8 object tracking code is already in 9 00:00:40,000 --> 00:00:44,000 Deepsort Object tracking code is already integrated with the YOLO V8 algorithm. 10 00:00:44,000 --> 00:00:50,000 So currently we have the object tracking code integrated with the YOLO V8 algorithm. 11 00:00:50,000 --> 00:00:54,000 We only need to add the object blurring code into it. 12 00:00:54,000 --> 00:00:54,000 Okay. 13 00:00:54,000 --> 00:00:59,000 So we will clone this github repo into our Google colab notebook over here. 14 00:01:00,000 --> 00:01:00,000 Okay. 15 00:01:00,000 --> 00:01:06,000 So I have created a GitHub code Google CoLab notebook for this project which you can see in front of 16 00:01:06,000 --> 00:01:07,000 your screen. 17 00:01:07,000 --> 00:01:14,000 So, so currently we have object tracking integrated with the YOLO V8 algorithm. 18 00:01:14,000 --> 00:01:17,000 We only need to add the object blurring code. 19 00:01:17,000 --> 00:01:23,000 So when we have the bounding box around the detected objects, we will blurred the object inside the 20 00:01:23,000 --> 00:01:24,000 bounding box. 21 00:01:24,000 --> 00:01:25,000 So this is our target. 22 00:01:25,000 --> 00:01:30,000 So first of all, I will clone this GitHub repo into our Google CoLab notebook. 23 00:01:31,000 --> 00:01:36,000 Before running the script, please make sure that you have selected the runtime as GPU. 24 00:01:36,000 --> 00:01:39,000 So now this GitHub repo is cloned into our Google CoLab notebook. 25 00:01:39,000 --> 00:01:41,000 And you can see this over here. 26 00:01:42,000 --> 00:01:47,000 Now we'll set this GitHub repo on GitHub repo as our current directory over here. 27 00:01:48,000 --> 00:01:48,000 Okay. 28 00:01:48,000 --> 00:01:49,000 So just ctrl v. 29 00:01:51,000 --> 00:01:51,000 So now. 30 00:01:52,000 --> 00:01:57,000 Now we need to install all the required libraries or the dependencies that are necessary to run this 31 00:01:57,000 --> 00:01:58,000 script successfully. 32 00:01:58,000 --> 00:02:04,000 So for this I will just run this cell so it will install all the required libraries that are required 33 00:02:04,000 --> 00:02:05,000 to run this script. 34 00:02:05,000 --> 00:02:11,000 So if you skip this step, then while doing the training validation or the prediction, you might face 35 00:02:11,000 --> 00:02:14,000 this error that the Hydra library is not installed. 36 00:02:14,000 --> 00:02:16,000 Matplotlib library is not installed. 37 00:02:16,000 --> 00:02:20,000 Seaborn Library is not installed, so it's necessary to run this step. 38 00:02:20,000 --> 00:02:24,000 So when you run this cell, you will see that requirements already satisfied, which means that some 39 00:02:24,000 --> 00:02:27,000 libraries are by default installed in the Google CoLab notebook. 40 00:02:27,000 --> 00:02:32,000 While some libraries are not by default installed the Google CoLab notebook. 41 00:02:32,000 --> 00:02:32,000 So it will be. 42 00:02:32,000 --> 00:02:40,000 So those library will be downloaded into the Google CoLab notebook like what you can say installed into 43 00:02:40,000 --> 00:02:41,000 the Google CoLab notebook. 44 00:02:41,000 --> 00:02:45,000 When you run this cell, like you can see that many of the libraries are not pre-installed. 45 00:02:45,000 --> 00:02:53,000 And when you run this cell, many of the libraries get installed, which include requests markdown Watchdog 46 00:02:53,000 --> 00:02:55,000 CD and all these libraries. 47 00:02:55,000 --> 00:02:58,000 So these libraries are not installed. 48 00:02:58,000 --> 00:03:03,000 So when we run this step, these libraries get installed while some libraries are by default installed 49 00:03:03,000 --> 00:03:05,000 in the Google CoLab notebook. 50 00:03:06,000 --> 00:03:08,000 So, uh, as we are doing. 51 00:03:10,000 --> 00:03:11,000 As we are doing the detection. 52 00:03:11,000 --> 00:03:14,000 So I will just navigate to the detection folder over here. 53 00:03:14,000 --> 00:03:16,000 So just give me a minute. 54 00:03:16,000 --> 00:03:17,000 Let's just refreshing. 55 00:03:17,000 --> 00:03:18,000 Okay. 56 00:03:18,000 --> 00:03:22,000 So if we see over here, just give me a minute, please. 57 00:03:22,000 --> 00:03:22,000 Okay. 58 00:03:22,000 --> 00:03:30,000 So as we see over here, if we go to Ultralytics, then we go to YOLO and we go to V8, then we go to 59 00:03:30,000 --> 00:03:31,000 detect. 60 00:03:31,000 --> 00:03:36,000 So we need to set this as our current directory because we are performing object detection and tracking. 61 00:03:36,000 --> 00:03:38,000 We are not doing segmentation. 62 00:03:38,000 --> 00:03:39,000 Okay. 63 00:03:39,000 --> 00:03:43,000 So just set this as your current directory and just run this cell. 64 00:03:45,000 --> 00:03:45,000 Okay. 65 00:03:45,000 --> 00:03:51,000 So as we are implementing object tracking using Deepsort, what we do in object tracking is we assign 66 00:03:51,000 --> 00:03:53,000 a unique ID to each of the detected object. 67 00:03:53,000 --> 00:03:58,000 This is the only function of object tracking to assign a unique ID to each of the detected objects. 68 00:03:58,000 --> 00:04:03,000 So to implement object tracking, I am using the deepsort algorithm, so we need to have the Deepsort 69 00:04:03,000 --> 00:04:07,000 files into our YOLO repository. 70 00:04:07,000 --> 00:04:11,000 So I'm just downloading the Deepsort files into my repository over here. 71 00:04:11,000 --> 00:04:17,000 So if you see over here you can see that deepsort zip file over here. 72 00:04:18,000 --> 00:04:19,000 Okay, you can. 73 00:04:19,000 --> 00:04:21,000 So we need to unzip this file first. 74 00:04:21,000 --> 00:04:27,000 So I'm just unzipping the deepsort file over here and you can see the unzipped file over here. 75 00:04:27,000 --> 00:04:32,000 I've already explained the YOLO V8 with deepsort object tracking code, so I will not discuss the YOLO 76 00:04:32,000 --> 00:04:36,000 V8 with deepsort object tracking code will only add the object planning code. 77 00:04:36,000 --> 00:04:38,000 If you want to learn this. 78 00:04:38,000 --> 00:04:40,000 I have made multiple video tutorials above. 79 00:04:40,000 --> 00:04:43,000 In the second section you must watch this videos. 80 00:04:44,000 --> 00:04:48,000 Okay, then I will download a sample video for testing from Google Drive. 81 00:04:48,000 --> 00:04:49,000 Okay. 82 00:04:50,000 --> 00:04:57,000 So first of all I will go to the predict.py file and I will add the object bearing code over here. 83 00:04:57,000 --> 00:04:59,000 So this will be a few lines of code. 84 00:04:59,000 --> 00:05:01,000 It will not be more than a few lines. 85 00:05:01,000 --> 00:05:06,000 So first of all, I will just right over here go to below. 86 00:05:06,000 --> 00:05:09,000 So I'm just scrolling down below and. 87 00:05:12,000 --> 00:05:15,000 Now, you can see here I've added the object blurring code. 88 00:05:15,000 --> 00:05:17,000 So let me explain you the code. 89 00:05:18,000 --> 00:05:25,000 In this step, we are just selecting the area where we have the area of the frame where we have the 90 00:05:25,000 --> 00:05:26,000 bounding boxes. 91 00:05:26,000 --> 00:05:30,000 Okay, so here I'm just passing the coordinates of the bounding boxes over here. 92 00:05:30,000 --> 00:05:35,000 I'm just passing the individual coordinates of the bounding boxes in the frame over here. 93 00:05:35,000 --> 00:05:40,000 So I will just crop the area first where we have the bounding boxes. 94 00:05:40,000 --> 00:05:41,000 Okay. 95 00:05:41,000 --> 00:05:43,000 Then I will plot the area. 96 00:05:43,000 --> 00:05:46,000 So then I will blurred those bounding boxes using cv2 dot blur. 97 00:05:46,000 --> 00:05:49,000 And here I've just set up blur ratio over here. 98 00:05:49,000 --> 00:05:50,000 Okay. 99 00:05:50,000 --> 00:05:58,000 Then I will just in the complete frame of the video, I will just add the blurring, add the blur code 100 00:05:58,000 --> 00:06:01,000 on the area where we have the bounding boxes. 101 00:06:01,000 --> 00:06:04,000 So I'm just blurring the detected object. 102 00:06:04,000 --> 00:06:06,000 So I have the bounding boxes around the detected object. 103 00:06:06,000 --> 00:06:09,000 So I'm just blurring the bounding boxes. 104 00:06:09,000 --> 00:06:09,000 Okay. 105 00:06:10,000 --> 00:06:17,000 Of the detected objects over here and then I'm just integrating this code into my original of frame 106 00:06:17,000 --> 00:06:18,000 of the video. 107 00:06:18,000 --> 00:06:21,000 So that is the three lines of code which I have added. 108 00:06:21,000 --> 00:06:25,000 So let's just run this cell and see what results do we get. 109 00:06:25,000 --> 00:06:30,000 So just again, checking the code, if there is any thing we can fix it truly. 110 00:06:30,000 --> 00:06:32,000 Okay, so that's fine. 111 00:06:32,000 --> 00:06:34,000 I think I have missed one thing. 112 00:06:34,000 --> 00:06:36,000 I think I need to define the import. 113 00:06:36,000 --> 00:06:37,000 Cv2 is done. 114 00:06:38,000 --> 00:06:39,000 Everything is done. 115 00:06:41,000 --> 00:06:45,000 Let's run this script and see what results do we get. 116 00:06:45,000 --> 00:06:50,000 If there is any error, we can fix it, but am hopeful that it will work fine. 117 00:06:50,000 --> 00:06:57,000 So it might take some time for to execute and it will process each of the frames separately. 118 00:06:57,000 --> 00:06:58,000 Okay. 119 00:06:58,000 --> 00:07:02,000 So the processing on the video has started over here. 120 00:07:02,000 --> 00:07:03,000 We can see it. 121 00:07:03,000 --> 00:07:07,000 And okay, so the processing on the video has started. 122 00:07:07,000 --> 00:07:09,000 So this will take some time. 123 00:07:09,000 --> 00:07:16,000 So either it has divided the complete video into 942 frames and it is doing the processing on each of 124 00:07:16,000 --> 00:07:18,000 the frames individually. 125 00:07:18,000 --> 00:07:23,000 Currently 147 frames are processed and currently 172 now. 126 00:07:23,000 --> 00:07:28,000 So it is doing the processing on each of the frame one by one. 127 00:07:28,000 --> 00:07:28,000 So. 128 00:07:29,000 --> 00:07:34,000 It will take some time and then we get the output video over here. 129 00:07:34,000 --> 00:07:38,000 So as this per execution complete, I will show you the output demo video. 130 00:07:39,000 --> 00:07:39,000 Okay. 131 00:07:39,000 --> 00:07:41,000 So this might take some time. 132 00:07:51,000 --> 00:07:51,000 Guys. 133 00:07:51,000 --> 00:07:56,000 The processing on the video is done and here we have the output video. 134 00:07:56,000 --> 00:08:01,000 Let me just download this video and show you what results do we get. 135 00:08:01,000 --> 00:08:07,000 Okay, So let me just play this output video and see whether we are able to do the object blurring or 136 00:08:07,000 --> 00:08:11,000 not, or whether we are able to do that, the detected objects or not. 137 00:08:11,000 --> 00:08:15,000 So you can see that we are able to blurred all the detected objects. 138 00:08:15,000 --> 00:08:20,000 You can see that where we have the bounding boxes on the those objects are being blurred. 139 00:08:20,000 --> 00:08:23,000 So let me go over here so you can see that. 140 00:08:24,000 --> 00:08:26,000 We are able to blur the detected objects. 141 00:08:26,000 --> 00:08:29,000 Plus we have also implemented object tracking. 142 00:08:29,000 --> 00:08:38,000 Like you can see that each object is being assigned a unique ID 10413211910916. 143 00:08:38,000 --> 00:08:41,000 Plus we have the tracking trails as well. 144 00:08:41,000 --> 00:08:48,000 So in this project we have implemented an integrated object blurring code with Deepsort object tracking, 145 00:08:48,000 --> 00:08:50,000 YOLO, V8 and YOLO V8. 146 00:08:50,000 --> 00:08:53,000 So that's all from this video tutorial. 147 00:08:53,000 --> 00:08:55,000 I hope you have learned something from this video tutorial. 148 00:08:55,000 --> 00:08:57,000 See you all in the next video. 149 00:08:57,000 --> 00:08:58,000 Till then, bye bye.