what toolbox functions can i use for shape and object recognition

Hi, i have a question about an assignment i have to do. In a assigment i need to do 3 things: I take a picture of a IC integrated circuit on a black piece of paper. Then i imread it into my program.

1 - Count the amount of pins on a IC (integrated circuit) - DONE with thresholding a bw image + dilation + labeling and regionprops. That is can sort of do.

2 - The program needs to recognize if its NOT a picture of a IC (integrated circuit). I know is asked but i dont know what toolbox function would me suited. I guess some toolbox function do shape / object recognition? but i cant determin what is best.

3 - The program need to check if all the pins are connected to the body of the ic. IC's always have a even number of pins so that shall be a if else construction.

I especially need help with 2.

code can be provided

Answer 1

[Himanshu]

Hi,

I see that you are trying to recognize if an image is not of an IC (integrated circuit) and need help identifying suitable toolbox functions for shape and object recognition.

I would suggest using the Computer Vision Toolbox which can provide functions for object detection and recognition. You can apply 'bagOfFeatures' to extract features from images for classification using machine learning.

You can utilize 'trainImageCategoryClassifier' to train a classifier to recognize different categories, including ICs. Consider using pre-trained networks (e.g., AlexNet, GoogLeNet) for more robust object recognition.

Please refer to the below documentations for more information.

1. Design and test computer vision systems using Computer Vision Toolbox: https://www.mathworks.com/help/vision/
2. Bag of visual words object: https://www.mathworks.com/help/vision/ref/bagoffeatures.html
3. Image Category Classification Using Bag of Features: https://www.mathworks.com/help/vision/ug/image-category-classification-using-bag-of-features.html
4. Train an image category classifier: https://www.mathworks.com/help/vision/ref/trainimagecategoryclassifier.html
5. Pretrained Deep Neural Networks: https://www.mathworks.com/help/deeplearning/ug/pretrained-convolutional-neural-networks.html

I hope this helps.