Article is from https://blog.csdn.net/qq_35451572/article/details/80249259
Article Directory
- 1. What is the target detection?
- 2. The object detection to solve the core problem
- 3. Target Detection learning resources
- 4. Target Detection latest developments
- reference
1. What is the target detection?
** ** target detection task is to find the target in the image all interested (object) to determine their location and size, is one of the core issues in the field of machine vision. Since various types of objects have a different appearance, shape, posture, plus interference light imaging elements, occlusion, object detection field has been the most challenging machine vision problems.
Computer vision, image recognition on four categories task:
Category -Classification: "? What is a" problem, that is what the target class given a picture or a video which contains judgment.
Positioning -Location: solve the "Where?" Question, that is, locate the position of the target.
Detection -Detection: solve the "What is Where??" Question, namely locate the position of the target and know what target Yes.
Segmentation -Segmentation: divided into instances (Instance-level) and the scene division (Scene-level), to solve the problem "of each pixel which belongs to the object or scene," the.
2. The object detection to solve the core problem
In addition to image classification, object detection to solve the core problem is:
1. Target may appear anywhere in the image.
2. Objectives There are various sizes.
3. The target may have different shapes.
If the target is defined by a rectangular frame, has a different aspect ratio of the rectangle. Since different aspect of the target, and therefore the sliding window + classic image scaling scheme common target detection cost solution is too high.
3. Target Detection learning resources
3.1 target detection papers, code cleanup
This site is a finishing below visual computing sites, top papers will be updated regularly and related codes recently published. And this is just a classification target detection, interested students can go look in other directions finishing papers and codes
https://handong1587.github.io/deep_learning/2015/10/09/object-detection.html
4. Target Detection latest developments
With the rapid development of computer hardware and depth of learning, object detection combined with the depth of learning, object detection also developed rapidly, this section will describe in detail (Update) in subsequent sections.
Target Detection (a) - a target detection Summary
Target detection (two) - Evaluation
Target Detection (iii) - R-CNN
Target Detection (D) - SPP-Net
Target Detection (V) - Fast R-CNN
Target detection (six) - RCNN Faster
Faster RCNN install and run Demo
Target detection (seven) - YOLO
Target detection (eight) - YOLO v2
YOLO v2 installation training test
darkflow installation testing
Target detection (nine) - YOLO v3
reference:
Depth study based on the target detection algorithm review -SigAI