You can use state-of-the-art object detectors such as Yolo3 and MaskRCNN. But you need to be aware that it will only detect objects which it has been trained on. In addition, it will give you bounding boxes (one step before the output of MaskRCNN or you can use Faster-RCNN) and from these boxes, you can determine the spatial location of the object with respect to the image.

You can use state-of-the-art object detectors such as Yolo3 and MaskRCNN. But you need to be aware that it will only detect objects which it has been trained on. In addition, it will give you bounding boxes (one step before the output of MaskRCNN or you can use Faster-RCNN) and from these boxes, you can determine the spatial location of the object with respect to the image.

Thanks for your answer Tahmid Chowdhury .

Dear Majid Farzaneh ,

Locate object in imageObject

Step 1: Read Images. Read the reference image containing the object of interest. Step 2: Detect Feature Points. Detect feature points in both images. ...

Step 3: Extract Feature Descriptors. ...

Step 4: Find Putative Point Matches. ...

Step 5: Locate the Object in the Scene Using Putative Matches. ...

Step 7: Detect Another Object. https://towardsdatascience.com/object-detection-with-10-lines-of-code-d6cb4d86f606 Regards,

Shafagat

Read about research on images you will find many researches with different topics

Thanks for answers Wafaa Alhashemi , Shafagat Mahmudova

if your image basically is georeferenced (geo-coded) you can find the coordinates of all objects in the real world.

if its not important the correct location of those in the world, you can write the number of x and y of image matrix. you can see this numbers by paint software or matlab indicator.

Thanks @Babak.

But I want to do that automatically.

The best way to find the location of an object on the image is to use thresholding, classification, and segmentation techniques.

Using Matlab can be a practical solution.

You can classify every pixel in your image, in this way you can classify a object. The best way for me is use Deep learning and neural network and model like unet.

Refer the following link. It may be helpful to you.

https://towardsdatascience.com/object-detection-with-10-lines-of-code-d6cb4d86f606

Object Detection in a Cluttered Scene Using Point Feature...

It depends on what kind of image are you going to localize and landmark if it's common and general object such as face, you can easily use dlib 68 or 194 point landmarks, but for the specific object such as blood cells you have to use the detection algorithm such as yoloV3, SSD or faster R-CNN it train them by your specific samples.