WEBVTT 00:00.810 --> 00:01.920 Hey guys welcome back. 00:01.920 --> 00:04.860 In this lecture let us talk about some important stuff. 00:04.860 --> 00:07.580 So now will we start playing with packages. 00:07.770 --> 00:13.230 Also the second thing is when we start something related to face or organization on how to play with 00:13.320 --> 00:14.340 images. 00:14.340 --> 00:16.170 So this is going to be a fun lecture. 00:16.310 --> 00:17.460 Let us get started. 00:17.970 --> 00:20.710 So in this lecture let us talk about open TV. 00:20.730 --> 00:26.070 So open TV is a open source library and it mainly deals with computer vision. 00:26.070 --> 00:27.360 You see that TV. 00:27.390 --> 00:30.570 It's an open source computer vision library. 00:30.600 --> 00:37.200 It was basically released for C++ in the year of 2000 and then it regularly get updated. 00:37.200 --> 00:44.940 We will be using open python which is a wrapper of original open CV that was written and C++ one of 00:44.940 --> 00:48.240 the best thing is they give regular updates. 00:48.240 --> 00:53.320 I will recommend you do a basic google search and read about open TV. 00:53.340 --> 00:57.110 It's a fun liability and also an interesting topic to read. 00:57.120 --> 01:04.320 Now let us come back to come on line and item so have I'll be performing a simple task for face recognized 01:04.320 --> 01:06.790 nation in the previous lecture I sure do. 01:06.800 --> 01:09.660 A simple demo while understanding our pay. 01:09.840 --> 01:15.500 In this lecture we'll be performing that the objective of this program is to detect faces. 01:15.540 --> 01:18.900 So we're trying to detect faces on particular images. 01:19.320 --> 01:21.880 I have downloaded four sample images. 01:21.900 --> 01:25.180 You can try to use them from the resource section. 01:25.180 --> 01:26.810 So this is Elon Musk. 01:26.850 --> 01:30.870 And the second one is also Ilan the third one is Mark Zuckerberg. 01:31.260 --> 01:32.980 And the fourth one is also Mark. 01:33.240 --> 01:39.660 Let me close these and get back to my script now before doing anything we need to keep three points 01:39.660 --> 01:43.740 in mind while performing any computer revision program. 01:43.740 --> 01:49.960 So the first thing is you need to match your python margin as well as the Latin really or the package 01:49.980 --> 01:51.270 which you are using. 01:51.270 --> 01:57.270 So far this example I would be using library called Open CV python which is the latest version with 01:57.270 --> 01:58.590 Python three point seven. 01:58.950 --> 02:03.640 So let me install that along with open CV. 02:03.640 --> 02:05.400 Our second requirement is no. 02:05.440 --> 02:09.160 Which is also a package in library and that is really important to deal with. 02:09.160 --> 02:14.590 That is it would be installed automatically while installing our open CV python. 02:14.890 --> 02:17.080 Now let me check my open CV virgin. 02:17.110 --> 02:18.380 I can actually see here. 02:18.400 --> 02:27.390 But let me show hey you can see my open CV Virgin is three point four point four point one nine. 02:27.610 --> 02:33.720 And if you want to install a specific virgin of anything you just need to use equals two and then define 02:33.730 --> 02:35.060 that was your name. 02:35.140 --> 02:36.120 Things looks fine. 02:36.130 --> 02:37.370 We have done our first. 02:37.390 --> 02:44.020 That is to install the second thing I want to talk about is we need to download a hot cascade find so 02:44.020 --> 02:50.740 hard cascade files are basically all classified that installed all the information for the object that 02:50.740 --> 02:51.880 we are going to retake. 02:51.880 --> 02:54.620 So in this example we are going to detect a phase. 02:54.640 --> 02:57.150 So there are several classifiers that are separate. 02:57.160 --> 03:03.490 It is there are several code modules that need to be checked and that is contain and are classified 03:03.640 --> 03:05.430 called Hard cascade classified. 03:05.920 --> 03:08.190 So I would be adding that in the resource. 03:08.230 --> 03:13.810 So this particular high cascade file is only for our face detection. 03:13.810 --> 03:19.150 If you have some idea diction there would be some separate classified if you want to detect something 03:19.150 --> 03:19.660 else. 03:19.720 --> 03:22.100 There would be different different classifiers. 03:22.180 --> 03:24.410 So if I open this one. 03:24.480 --> 03:29.580 So this is some more defined looks like it's quite long and it contains a lot of code. 03:29.820 --> 03:31.200 So I would not mess with this. 03:31.200 --> 03:32.690 It is an excel file. 03:33.000 --> 03:36.220 And it is basically only for our face to addiction. 03:36.300 --> 03:37.800 Don't forget to download that file.