Add img_hash module

[mdeleau]

Hello,

First of all, thank you for this project which allows me to kickstart an old project idea of mine. And I think it will help a lot of other developers.

I needed the img_hash module so I took the liberty to implement it.
I am a beginner in flutter and this is my first contribution to open source so do not hesitate to point at anything wrong in what I did.

I tried to stay as close as possible to the gocv implementation but I think some things can be improved, especially the tests.

[rainyl]

Great! Thanks for you contributions, I will take a look at it recently.

BTW, actually, the current implementation is not robust so I am working on rewriting the implementations of gocv, including adding the ability to catch exceptions and avoid using bare pointers, this will make the exceptions more friently (#13 ), and I am trying to make the APIs more concise and similar to opencv-python, these are breaking changes and may be released in v1.0.0.

I will push recent commits to catch-exceptions branch, you can take a look at it and change the current implementation of img_hash to the new styles if you have more time.

[rainyl]

Minor please consider:

The current implementation is "static", the hash class instances will be freed after computation, maybe we can do someting like

opencv_dart/src/core/core.h

Line 95 in 76c542d

CVD_TYPEDEF(cv::RNG, RNG)

and provide operation interfaces like

opencv_dart/src/core/core.h

Lines 518 to 531 in 76c542d

	CvStatus Rng_New(RNG *rval);
	CvStatus Rng_NewWithState(uint64_t state, RNG *rval);
	void Rng_Close(RNG *rng);
	CvStatus TheRNG(RNG *rval);
	CvStatus SetRNGSeed(int seed);
	CvStatus RNG_Fill(RNG rng, Mat mat, int distType, double a, double b, bool saturateRange);
	CvStatus RNG_Gaussian(RNG rng, double sigma, double *rval);
	CvStatus RNG_Uniform(RNG rng, int a, int b, int *rval);
	CvStatus RNG_UniformDouble(RNG rng, double a, double b, double *rval);
	CvStatus RNG_Next(RNG rng, uint32_t *rval);
	CvStatus RandN(Mat mat, Scalar mean, Scalar stddev);
	CvStatus RandShuffle(Mat mat);
	CvStatus RandShuffleWithParams(Mat mat, double iterFactor, RNG rng);
	CvStatus RandU(Mat mat, Scalar low, Scalar high);

This can also solve: https://github.com/mdeleau/opencv_dart/blob/ccba658be755ddd7d5551daea9691070cc961347/lib/src/contrib/img_hash.dart#L93

But for current status it is fine and feasible, like I said, in the future we will refactor this.

[rainyl]

I reviewed it briefly and it's quiet great, no much to change and I'm going to merge it in to main branch.

I will be grateful if you are willing to refactor it for #13 , to do this please open PRs into catch-exceptions branch, but if you have no time, don't worry, I will finish it in the future.

[mdeleau]

Hello, thank you for the review. I do not think I will have the time to refactor it during the next weeks. But I will try to refactor it as soon as I can.