7

I don't think pitch information is relevant for what you want to do. The variation of pitch during speech is known as intonation, and can convey emotions, indicate if a sentence is a question etc. However, there is no universal rule as to how pitch variation patterns are mapped to meaning - this is quite language dependent ; and some languages sound "...


7

What happens to the video: In the most simple implementation, which is suitable for large integer fast forward rates, only 1 frame out of N, where N is the fast forward rate is played. For example, with a normal play rate at 30 FPS, frames 0 to 29 are played in sequence over a duration of one second. In fast forward with a 3x rate, frames 0, 3, 6 ... 87 are ...


5

Monocular video is taken using a single camera. Stereo video is taken using two cameras side-by-side, producing two video streams. That's pretty much it. A stereo video pair allows depth measurement, so you can find out how far away things are. However, this is not an easy task for complex scenes.


4

Averaging may not be the best option, due to motion compensation in video frames: one number (say $8$) may be extrapolated for another (say $3$) on a different JPEG block. It might be interesting to dig into the compressed video format, check motion vectors, but this requires a little bit of hacking. Some software which can help: Amped FIVE (used by the ...


3

To OCR or not? If the timer is about $h$ pixels height, you could segment each frame into lines of height $h/2$, run an OCR on each line, and scan the result with a regexp for anything that looks like a timer, eg. \d+[\.:,;-]\d+[\.:,;-]. This gives you $x=time_{real}$ vs. $y=time_{video}$ data pairs. Sure, most of them are wrong, but imposing the ...


3

Since you already know your background image beforehand, it should be simple. I done many background subtraction before. Here is an example of removal of background by comparison. if (newBitmap.Width == backgroundBitmap.Width && newBitmap.Height == backgroundBitmap.Height) { for(int x = 0; x < newBitmap.Width; x++) { ...


3

Bit rate and data rate are somewhat ambiguous, and their exact definitions vary from one field or application to another. Bit rate is often used to measure the information rate, since information is measured in bits. Sometimes it is used to measure the number of actual 1s and 0s transmitted per second. When the information is compressed (source-coded), the ...


3

[EDIT: addition of a March 2017 preprint] Deep learning already has many applications in video, like enhancement (Deep Convolutional Neural Network for Decompressed Video Enhancement) or semantic analysis. Recently, there have been some announcements related to video compression, for instance: Generative Compression (Preprint, March 2017) Traditional ...


3

The frame is 2D Grid which samples the continuous world in Discrete way. Leave alone video for a second. Let's talk about taking a photo of a white paper laid on a black surface. The photo is taken from above and let's assume there is no lens issues. Do you expect the end of the paper to end at edge of a pixel of the camera perfectly? It won't. The end ...


3

For those classic Video Processing operations there is nothing better than the Plug In's of AviSynth. Specifically for De Flickering look at: LMFlicker. ReduceFlicker. DeFlicker. The source code of each is available so you can get inspiration from them. All of them basically works on smoothing the Luma Data along the temporal dimension.


2

From an algorithm perspective, subtraction is the simplest method. Run a feature detection algorithm (like SIFT or SURF) on both images to align them Apply any enhancement operations at this stage, such as lens distortion removal, lighting correction, or others. Do a simple subtraction of the background image from the test image. Perform a blur or median ...


2

This is clearly not an easy task. The problem is, if you want a more-or-less accurate count, then you need to turn to advanced algorithms (and maybe use 2 cameras, or a kinect). If you can't afford to take this path, then you need to try simpler options. Personally, I would try the following: detecting skin pixels, segmenting the image with respect to ...


2

There are many ways to skin this cat. I'll explain one way that I know is used in industry. Create candidate frames algorithmically using the image features; e.g., the histogram, sharpness, presence of humans, faces, and anything else you can think of. Serve different candidates to different users and see which ones are popular You can stop once the ...


2

Scenes generally exhibit a fade-to-black transition. You could capture those frames with simple image processing tools, which are found in many libraries, such as OpenCV. If you want to rely on the change of the mis-an-scene to be robust against sudden changes, then of cours algorithms for detecting temporal differences is more appropriate. For that, you ...


2

One way to do it is to solve a MAP problem of the up scaled video and using the High Resolution images as a prior. Try looking at the articles - Super Resolution MAP.


2

I think Technically speaking the only way to detect when a frame ends is to see where the next one begins. That means, you should search for the next start prefix code 0x000001. If you go to the previous data before this (which in the general case is the byte alignment syntax elements) and add your RS codes, a compliant decoder should ignore all that until ...


2

There is a company named Signalscape who has a line of products called StarWitness (http://www.starwitnessinfo.com/) that do forensic video analysis that are available to law enforcement agencies (only). I will try to get a hold of the director of the StarWitness group to determine if there is some course of action you can take that may assist you in ...


2

Usually, Salt and Pepper Noise is done using Median Filter. There are variations of the Non Local Means to deal with Salt and Pepper but it might be overkill so I'd start with the above. The tricky thing is to take advantage of having a Video and not just a static single image. Hence you can also apply the filter in a temporal form. If you use small kernel ...


2

I've used in the past Julie Delon's flickering removal (pdf) approach which is based on midway image equalization. It's not hard to implement, and the parameters are easy enough to tune. Since it seems that the brightness change impacts the whole frame, you probably don't need to use the patch oriented approach and can go for a pair-wise midway throughout ...


2

When upsampling the number video frames to allow playing a video in slow motion, rather than for a frame rate increase for smoothness, you will likely need to modify the audio using a time-pitch stretching/shifting algorithm to stretch the audio out (increase the number of samples) for a longer play duration (to match the increase in slow motion video ...


2

No. The MPEG4 video codec allows for a large ranges (easily a dynamic range of 50 for "useful" quality) of compression ratios for the same raw video material. The same applies to audio compressors. And of course, there's simply video material that is easy to compress, so using the same type of quantization, motion prediction and frame structure, some ...


1

The ffmpeg implementation of the HEVC decoding follows this formula instead: $ |p_2,_0 - 2p_1,_0+p_0,_0|+|p_2,_3 - 2p_1,_3+p_0,_3|+|q_2,_0 - 2q_1,_0+q_0,_0|+|q_2,_3 - 2q_1,_3+q_0,_3|<\beta \ (1)$ This appears in the hevc_deblock.asm file. ;compare pcmpgtw m15, m13, m14 movmskps r13, m15 ;filtering mask 0d0 + 0d3 < beta0 (bit 2 ...


1

Reduction of Noise in a video depends on what type of noise is present in the video. The noise can be reduced by filters which are all present [There are software for this] and many type of filters such as median ,mean,wiener filter. And The noise model can be build to add noise to the video there are many matlab code as well as software also there. Gaussian ...


1

Implement a low pass 2D filter that will reduce greatly the noise. Following an example that you could complete yourself. You could subtract the filter bias to keep an average of 0 "addition" by the filter, controlling the dynamic range. Example of filter 1/16 1/8 1/4 1/8 1/16 1/16 1/8 1/4 1/2 1/4 1/8 1/16 1/16 1/8 1/4 ...


1

What about computing the sum of squared differences or the correlation between the current frame and the previous frame? For fast moving scenes, you could compute phase correlation via Discrete Fourier Transform, but I think that computation would be too expensive for video processing, especially if there's a real-time requirement.


1

Probably you could try optical flow with some key-points not pixels if your purpose is to estimate motion. I thing meaningful thing would count average values of keypoints, overcomes some threshhold or filter (to eliminate barelly moving keypoints). However there would be some problems with having constant amount of meaningfull keypoints (keypoints tend to ...


1

Yes you can re-represent the 2d signal as a 1-d signal and use signal processing algorithms on it. In doing this you can lose some of benefits of representing the data this way. Eg. If you were looking at compression of a mostly continuous 2-d dataset you may get benefits by treating the data as 2-dimensional and compressing in each dimension. Representing ...


1

Naive question: is it very common for the relationship between the two sequences of timestamps (the "warping function") to be something else than a linear function? I assume it'll be almost always linear, in which case you just need to match two points (say the first and the last non-black frame), and then you can interpolate between that. If you want to do ...


1

To compute an offset between the two signals in the first image, I think that phase-correlation will give you much better results, than simply the basic correlation. You can reuse your fft code but have to normalize your fft coefficients to unit magnitude prior to your inverse-fft, so the phase-correlation is based only on phase information and is ...


1

There is an open-source solution to be used with Python: Pyradi: an open-source toolkit for infrared calculation and data processing, SPIE Proceedings, Edinburgh, 24-27 September, C. J. Willers, et al. The PyRadi toolkit (version 1.1.0, 2016-11-20) is a Python toolkit to perform optical and infrared computational radiometry (flux flow) calculations: The ...


Only top voted, non community-wiki answers of a minimum length are eligible