Every industry has an algorithm part, statistics has statistical algorithms, control has control algorithms, and image processing has image processing algorithms. In many traditional industries, algorithm is not an independent position, but is the responsibility of R&D engineers. Today, the editor will take you to understand roughly what an algorithm engineer does? Let’s read on.

1. Image processing, especially the image processing algorithm based on OpenCV. Generally, products include beautification, filters, etc. Children who especially like to recruit this area have been used for deep learning in the past year or two. replacement trend. Recently, Google released arcore, which has allowed many small companies to produce some very effective head-changing applications.

2. Computer graphics, which can be considered a large category, mainly involves image rendering aspects such as graphics rendering algorithms, ray tracing algorithms, and three-dimensional image reconstruction. This direction is not only for 3D engine and game development, but also for many industries that need to be related to CAD, which will involve model and optimization algorithm design in this field.

3. The field of VR and AR involves video tracking, SLAM, raytracing, geometric projection, etc. It is actually a comprehensive field. Currently, it is mainly focused on computer vision.

4. Medical image processing, three-dimensional image reconstruction, used in B-ultrasound and CT imaging, this is in the medical direction.

5. Communication baseband signal processing and network optimization algorithms are actually in decline. After all, small companies do not have the cost to implement advanced algorithms.

6. Audio filtering is used in HiFi products, such as car audio and mobile phone manufacturers. The circle is actually quite small.

7. Control algorithm, adaptive filtering algorithm, used in the mechanical field, such as robot arm stroke control and stability.

8. Finite element algorithm has valuable applications in radar, machinery, electromagnetics, and clothing design.

9. Signal processing, such as interpolation, spectrum analysis, blind signal separation, compressed sensing, most applications of the Internet of Things will involve this area.

There are generally two reasons why the Internet and software industries separate algorithms into an independent position. First, low-level software engineers do not understand algorithms, or to put it more simply, they do not understand mathematics. All work involving models and calculation formulas must be done by professionals. Second, from the perspective of production efficiency, many junior algorithm engineers do not have a good software engineering background. To put it simply, they cannot write code and can only write matlab. There is no way for such engineers to directly put their work into production, so they need to put their work into production. Work and production links are isolated. To sum up, this is what the editor will share with you today. I hope it can help you.