Crop identified
Budget: $30 – $250 USD
I need a Mac OS application, written in Swift, to take a user selected folder of images and extract all animal objects identified in those photos as individually cropped images into a separate folder.
1. User asked for a folder of photos to analyze
2. User is asked for a folder for where extracted images should be saved
3. App reads in each photo one at a time
4. App uses Apple's Vision api/library to find all animals in each photo
5. App saves each of the animals individually found in each photo as a cropped JPG using the naming convention <original file name> - <classification(s) of animal> - <sequence number like 001...>.jpg using an image quality of 100. The sequence number should just be an auto incrementing value starting at 001 of each photo analyzed so that if there are multiple birds found in a photo they would each have unique sequence numbers like 001, 002, 003 to prevent file name conflicts.
Attached is a copy of Apple's sample code for doing image classification that already does much of what I'm looking for. It asks the user for a folder of photos, it then analyzes the images and returns classifications of objects found in the photos and all the photos that have that classification. I don't care about viewing the results like that, rather, I would just like the app to automatically take those classifications it finds for each photo and save a cropped image for each of those classified objects found in each image.
I have also attached a few sample photos and have done manually cropping on them based on what I hope the app will be able to do, to demonstrated the intended output.
Finally, if there are other publicly available ML libraries that can be utilized in the Mac OS App to assist in better finding the animal (although birds are my primary focus) outlines to get coordinates for cropping, I am absolutely open to that as Apple's Vision system may not be the most effective tool to do this.
WHEN SUBMITTING INTEREST IN THIS JOB PLEASE SUPPLY:
1. A brief statement about your experience developing vision based apps/utilities on the Mac or iOS using Swift and Xcode.
2. Just a few brief examples, IF AVAILABLE and you are authorized by that client to share information about that project, of similar work if you feel any may be helpful to demonstrate your prior experience.
3. Questions you have that are necessary for you to be able to adequately provide a time/effort estimate for this job.
Thank you for considering this job. I promise to answer questions as quickly as I am able. I promise to be open to suggestions and recommendations based on your professional experience. I promise to pay on-time, as agreed to. I promise to provide fair feedback and a review of your services at the completion of the project.
1. User asked for a folder of photos to analyze
2. User is asked for a folder for where extracted images should be saved
3. App reads in each photo one at a time
4. App uses Apple's Vision api/library to find all animals in each photo
5. App saves each of the animals individually found in each photo as a cropped JPG using the naming convention <original file name> - <classification(s) of animal> - <sequence number like 001...>.jpg using an image quality of 100. The sequence number should just be an auto incrementing value starting at 001 of each photo analyzed so that if there are multiple birds found in a photo they would each have unique sequence numbers like 001, 002, 003 to prevent file name conflicts.
Attached is a copy of Apple's sample code for doing image classification that already does much of what I'm looking for. It asks the user for a folder of photos, it then analyzes the images and returns classifications of objects found in the photos and all the photos that have that classification. I don't care about viewing the results like that, rather, I would just like the app to automatically take those classifications it finds for each photo and save a cropped image for each of those classified objects found in each image.
I have also attached a few sample photos and have done manually cropping on them based on what I hope the app will be able to do, to demonstrated the intended output.
Finally, if there are other publicly available ML libraries that can be utilized in the Mac OS App to assist in better finding the animal (although birds are my primary focus) outlines to get coordinates for cropping, I am absolutely open to that as Apple's Vision system may not be the most effective tool to do this.
WHEN SUBMITTING INTEREST IN THIS JOB PLEASE SUPPLY:
1. A brief statement about your experience developing vision based apps/utilities on the Mac or iOS using Swift and Xcode.
2. Just a few brief examples, IF AVAILABLE and you are authorized by that client to share information about that project, of similar work if you feel any may be helpful to demonstrate your prior experience.
3. Questions you have that are necessary for you to be able to adequately provide a time/effort estimate for this job.
Thank you for considering this job. I promise to answer questions as quickly as I am able. I promise to be open to suggestions and recommendations based on your professional experience. I promise to pay on-time, as agreed to. I promise to provide fair feedback and a review of your services at the completion of the project.