Monday

We had a meeting with Sven who walked us through a series of steps on applying Sobel filters to images which will enable us to calculate the derivatives from an image by creating a binary edge map.

We also discussed how we can segment various sections of the image for the algorithm to single out IC’s in order to give more accuracy in the matching.

Tuesday

Our two tasks after our meeting with Sven were to research on two methods better than blob detectors; Hough and Distance Transform. Results of these methods were far-fetched from what we expected and didn’t prove useful. Nonetheless, we learned something new.

Wednesday

We spent the rest of the day researching and searching for sample code which will help us more in the segmentation processes. Gerald made significant findings in terms of making advancements in the project by finding an article which helped him write a program which basically identifies the IC’s in the image, crop them out, and generate a new image file for it.

Friday

We met with Dr. Crandall to show him the results, and he helped brainstorm with us to see how best we could make our code more efficient. He wanted us to try and see if we can use just the texts on the ICs to help make the matching even more accurate. To do so, he recommended us to use Computer Recognition programs that companies offer online such as Rekognition by Amazon, Google Vision, or Tesseract.

This upcoming week, we will play around with these Computer Recognition programs and try to clean up our file space. From there, we hope to get a range of accuracy for the matching algorithm we currently have been using.