How can I make this features extraction code to detect and count

1 回表示 (過去 30 日間)
hajer jon
hajer jon 2019 年 8 月 21 日
回答済み: Prabhan Purwar 2019 年 8 月 27 日
Hi
i need to make this code to detect and track people it extract feaures and combine it i an array
pointrgb=detectSURFFeatures(gryI2,'NumOctaves',1,'NumScaleLevels',3,'MetricThreshold',5000);
[features22, valid_points22] = extractFeatures((gryI2),pointrgb);
[reg,dd] = detectMSERFeatures(rgb2gray(I2),'ThresholdDelta',1,'MaxAreaVariation',1);
[features33, valid_points33] = extractFeatures((gryI2),reg);
comfeat2{k,:}=[features22(:,:);features33(:,:)];
valid_pointscom{k,:}=vertcat(valid_points33,valid_points22);

回答 (1 件)

Prabhan Purwar
Prabhan Purwar 2019 年 8 月 27 日
Hey,
Detection and Tracking are two interdependent processes.
Object Detection and Recognition Code based upon Point Feature Matching can be downloaded from the link attached.
Tracking of an object can be performed using the following links.
Note: Ignore the Augmented Reality (AR) part in the video, object detection and tracking based upon Point Feature Matching.
Although I would suggest you making use of more advanced detection and tracking techniques, such as Motion-Based Multiple Object Tracking, Face Detection and Tracking Using CAMShift, Kanade-Lucas-Tomasi (KLT) point tracking algorithm, Kalman Filtering to predict the location of a moving object and others.
Link to some of the popular techniques are listed below:

カテゴリ

Help Center および File ExchangeComputer Vision with Simulink についてさらに検索

製品


リリース

R2018a

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by