rgbx | hdr to rgba8 encoding/decoding tool | Computer Vision library

Problem

Basically I have this code to streaming the desktop screen. The problem is that when I will try to resize the screen of server (server will receive the images) the image stay cutted/distorted

My try of resize the window:

What is supposed to be (simulation):

Question

Which alteration is need to resize correctly the window of image streaming?

Thanks in advance.

Server

I've been having this issue for a while and I've ignored it for just as long, but at this point, the fact that it won't go away is making me see that it's a real issue. Whenever I load an image onto an object, the colors separate into bands like this:

Color separation in render

This is the image that I'm using:

What the image should look like

I'm not entirely sure if this is correct, but I've noticed that it only seems to happen on images with complexity. When I have an image that's a single color, it works fine, but textures and photographs don't seem to ever work. Also, note that I tried converting the original image to different file-types to see if that would help with no difference in results.

the functions that I'm using to load the image:

I'm running a neural network on a live stream of screenshots 800x600 in size. Since I was only getting about 3fps, I did some troubleshooting and found out how much time is approximately spent on each step:

• Screenshot: 12ms
• Image processing: 280ms
• Object detection and box visualisation: 16ms
• Displaying image: 0.5ms

I'm using mss for taking the screenshots (documentation).

Here's the code without the object detection part:

I am creating a 512x512 raster image in a byte string in "RGBX" fomat (from a memory-mapped device), and cannot get it to display in a label image. The image displays just fine with show(). I just need to quickly transfer the byte data directly to the image input for a button or label, is that possible?

I have tried converting the image with .convert, doesn't do RGB, PhotoImage wants a string variable only, base64.b64encode() spawns an unkillable zombie on my machine. I have tried to make the image object 'static' in the demo, I think most answers to similar problems point to making the image stay around to be displayed. An image I load by opening any file displays fine with the label or button image methods. io.BytesIO doesn't support the image into it.

I'm trying to grab video from a window using ximagesrc and scale it to a certain size before encoding in H.264 and streaming with RTP to another machine. I implemented my pipeline in the C API and it works fine unless I add a videoscale element with capsfilter.

Specifically, I have a problem understanding how to use the videoscale element correctly and how to link it with a videoconvert element programmatically. The function gst_element_link_filtered returns false when I try to connect the videoconvert and videoscale element using a capsfilter for scaling to the resolution I want.

My code looks as follows:

I haven't written many Metal kernel shaders yet; here's a fledgling "fade" shader between two RGBX-32 images, using a tween value of 0.0 to 1.0 between inBuffer1 (0.0) to inBuffer2 (1.0).

Is there something I'm missing here? Something strikes me that this may be terribly inefficient.

My first inkling is to attempt to do subtraction and multiplication using the vector data types (eg. char4) thinking that might be better, but the results of this are certainly undefined (as some components will be negative).

Also, is there some advantage to using MTLTexture versus MTLBuffer objects as I've done?

Currently I'm developing a jpeg encoder for a university project. The image has fixed size and the encoder uses fixed quantization and Huffman tables for baseline process. First, my code reads 32-bit RGBx values from SDRAM, converts to YCbCr color space, normalizes each channel to 0 and writes back. Then, it starts doing DCT on 8x8 blocks and writes entropy encoded data to SDRAM. This process is done using C and then a Python code creates a file with appropriate JFIF markers and the entropy encoded data. Last of all, OS default jpeg decoder is used to view the image by simple double-clicking on it.

My code works with 8x8, 8x16 and 16x8 images but not with 16x16 nor the actual size of the image used in project. Below you may see 16x16 example.

16x16 Input

16x16 Output

However, on stackoverflow it seems different than compared to my OS' default decoder. Below is how it looks like on macOS Preview application.

I believe my problem is due to either the markers in JFIF or some kind of an algorithm error.

I would be very glad if anyone with experience in jpeg can help me out.

Kind regards