Add the UnsharpMask Effect

effect.c

@@ -5,3 +5,9 @@ convolveImage(Image *image, void *data, ExceptionInfo *ex) {
     ConvolveData *d = data;
     return ConvolveImage(image, d->order, d->kernel, ex);
 }
+
+Image *
+unsharpMaskImage(Image *image, void *data, ExceptionInfo *ex) {
+    UnsharpMaskData *d = data;
+    return UnsharpMaskImage(image, d->radius, d->sigma, d->amount, d->threshold ,ex);
+}

effect.go

@@ -32,3 +32,17 @@ func (im *Image) Convolve(order int, kernel []float64) (*Image, error) {
 	data.kernel = (*C.double)(unsafe.Pointer(&kernel[0]))
 	return im.applyDataFunc("convolving", C.ImageDataFunc(C.convolveImage), &data)
 }
+
+// UnsharpMask sharpens one or more image channels. We convolve the image
+// with a Gaussian operator of the given radius and standard deviation (sigma).
+// For reasonable results, radius should be larger than sigma. Use a radius of 0 and
+// UnsharpMaskImage selects a suitable radius for you.
+func (im *Image) UnsharpMask(radius float64, sigma float64, amount float64, threshold float64) (*Image, error) {
+	var data C.UnsharpMaskData
+	data.radius = C.double(radius)
+	data.sigma = C.double(sigma)
+	data.amount = C.double(amount)
+	data.threshold = C.double(threshold)
+
+	return im.applyDataFunc("unsharp_mask", C.ImageDataFunc(C.unsharpMaskImage), &data)
+}

effect.h

@@ -8,6 +8,14 @@ typedef struct {
     double *kernel;
 } ConvolveData;
 
+typedef struct {
+	double radius;
+	double sigma;
+	double amount;
+	double threshold;
+} UnsharpMaskData;
+
 Image * convolveImage(Image *image, void *data, ExceptionInfo *ex);
+Image * unsharpMaskImage(Image *image, void *data, ExceptionInfo *ex);
 
 #endif