WEBVTT 00:00.120 --> 00:07.880 We converted the normalized center to pixel coordinates, and we succeeded in drawing the bounding box 00:08.160 --> 00:15.400 and applying clamping boundaries to contour to ensure the rectangle stays within the visible screen. 00:15.440 --> 00:17.720 Now let's scroll up. 00:17.840 --> 00:18.160 Here. 00:18.160 --> 00:24.880 We created the results, which is a mutable list of detection object and we need to return it. 00:24.880 --> 00:31.120 So for this purpose we need to return the results dot sorted by ascending. 00:31.440 --> 00:35.240 But our results mutable list is empty. 00:35.480 --> 00:38.640 So here results dot add. 00:38.680 --> 00:42.840 We need to add this detection object detection object. 00:42.840 --> 00:49.280 And if we open the detection object we need to pass three parameters score, label and bounding box. 00:49.600 --> 00:57.360 The first uh the first is the score, the second is label and the third is the bounding box. 00:57.400 --> 00:57.840 Okay. 00:58.200 --> 01:01.320 This is the bounding box which is the rectangle okay. 01:01.600 --> 01:03.000 This is the results dot. 01:03.040 --> 01:08.420 Add the detection object and return results dot sorted by ascending. 01:08.540 --> 01:17.980 Let's try to run our application and see if there is any missing, uh, functions, any missing, uh, 01:18.220 --> 01:19.860 things we need to do. 01:19.900 --> 01:22.260 For example, this is the tensor image. 01:22.460 --> 01:26.860 Um, we missed to add it, but let me see. 01:26.900 --> 01:30.060 Let me run at first and see if there is any errors. 01:30.100 --> 01:35.460 This is our app working fine but it's not detecting any objects. 01:35.460 --> 01:38.780 So we need to check our codes. 01:38.820 --> 01:43.220 Let me scroll to the tensor image here. 01:43.340 --> 01:48.140 I've used the input buffer instead of using the tensor image. 01:48.140 --> 01:49.780 So input. 01:50.020 --> 01:55.540 Let me check the tensor image the tensor image dot buffer. 01:55.820 --> 02:01.860 So change the input buffer in to tensor image dot buffer. 02:02.140 --> 02:03.500 Let me run again. 02:03.540 --> 02:08.410 Here we have our application and it's detecting the battery. 02:08.450 --> 02:10.050 Let me run it again. 02:10.090 --> 02:10.850 Here we go. 02:10.890 --> 02:16.290 You see, guys, we are detecting the batteries, but our application is crashing. 02:16.570 --> 02:19.850 Let me check the the logcat. 02:19.890 --> 02:22.170 And by the way, congratulations, guys. 02:22.170 --> 02:23.930 This is a complex application. 02:23.930 --> 02:27.770 We are making our custom object detection application. 02:27.810 --> 02:36.250 Okay, if we check here we have this error array index out of bounds exception links. 02:36.250 --> 02:38.650 Fix six and index six. 02:38.810 --> 02:42.450 Array list dot get and drop boxes Lambda. 02:42.490 --> 02:43.970 Let me check that. 02:43.970 --> 02:48.290 Drop boxes dot key file and the error here. 02:48.290 --> 02:52.570 Inside draw a path color list. 02:52.570 --> 02:54.770 Let me check the color path. 02:55.090 --> 02:56.730 It's a list of colors. 02:56.770 --> 03:01.810 Go to to that list of colors that we passed for the drop boxes. 03:01.850 --> 03:09.090 I think it's in the object detector, so scroll up. 03:09.250 --> 03:13.950 Search for search for um path color list. 03:13.950 --> 03:20.950 So ctrl f path color list and apply at index I to the RGB. 03:20.990 --> 03:24.350 So maybe there is many detections. 03:24.350 --> 03:34.990 If you see guys there are a lot of detections and overlap detections um, on uh on the screen and it's 03:35.150 --> 03:39.310 uh it's not for example checking for the colors. 03:39.310 --> 03:40.710 We have six colors. 03:40.710 --> 03:41.990 Let me check guys. 03:42.150 --> 03:44.990 So open that detection screen. 03:45.150 --> 03:51.470 And here we have six parameters green, cyan, red, yellow magenta and black. 03:51.830 --> 03:55.070 And if you notice that the black box is drawn. 03:55.070 --> 03:59.830 So we reached the maximum size of this color list. 03:59.830 --> 04:05.190 Let me add the new colors in order to solve the problem. 04:05.190 --> 04:15.740 Also we can uh right here in order to prevent any errors, we can make a modulus or repeating the colors 04:15.780 --> 04:16.340 again. 04:16.580 --> 04:21.540 Okay, but since for for now we've added four more colors. 04:21.540 --> 04:28.060 When you reach white color, then the next detected one will make an error. 04:28.340 --> 04:31.140 Okay, let me test this application. 04:31.140 --> 04:37.380 And notice if if this the the real problem here is the detection. 04:37.380 --> 04:38.780 And it's working fine. 04:38.820 --> 04:42.740 Now you see guys everything is working fine. 04:42.740 --> 04:44.500 And congratulations. 04:44.700 --> 04:47.460 This is our lovely application. 04:47.500 --> 04:48.940 Those are two batteries. 04:48.940 --> 04:50.300 Let me test okay. 04:50.340 --> 04:51.540 You see guys, we are. 04:51.740 --> 04:54.940 We are detecting two batteries. 04:55.340 --> 04:56.340 Let me see. 04:56.780 --> 05:00.620 And this is what we trained our model, uh, for. 05:00.900 --> 05:04.540 So this is our lovely application. 05:04.580 --> 05:05.140 Okay. 05:05.420 --> 05:10.420 So congratulations, guys, and thank you for following me.