@niivue/nv-ext-image-processing 1.0.0-rc.3 → 1.0.0-rc.4

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@niivue/nv-ext-image-processing",
-  "version": "1.0.0-rc.3",
+  "version": "1.0.0-rc.4",
   "type": "module",
   "description": "Image processing extensions for NiiVue (Otsu thresholding, haze removal, etc.)",
   "main": "./dist/nv-ext-image-processing.js",
@@ -16,10 +16,10 @@
     "dist"
   ],
   "peerDependencies": {
-    "@niivue/niivue": "workspace:*"
+    "@niivue/niivue": "1.0.0-rc.4"
   },
   "devDependencies": {
-    "@niivue/niivue": "workspace:*",
+    "@niivue/niivue": "1.0.0-rc.4",
     "typescript": "~5.8.3",
     "vite": "^6.3.2",
     "vite-plugin-dts": "^4.5.4"
@@ -31,5 +31,8 @@
   "dependencies": {
     "gl-matrix": "^3.4.4",
     "nifti-reader-js": "^0.8.0"
+  },
+  "publishConfig": {
+    "access": "public"
   }
 }

package/dist/index.d.ts

@@ -1,8 +0,0 @@
-/**
- * @niivue/nv-ext-image-processing
- *
- * Image processing extensions for NiiVue.
- * Each export is a VolumeTransform that can be registered with
- * `ctx.registerVolumeTransform()` via the NiiVue extension context.
- */
-export { conform, connectedLabel, otsu, removeHaze } from './transforms';

package/dist/nv-ext-image-processing.js

@@ -1,154 +0,0 @@
-import { NVWorker as u } from "@niivue/niivue";
-const f = '(function(){"use strict";var T=typeof Float32Array<"u"?Float32Array:Array;function B(){var n=new T(16);return T!=Float32Array&&(n[1]=0,n[2]=0,n[3]=0,n[4]=0,n[6]=0,n[7]=0,n[8]=0,n[9]=0,n[11]=0,n[12]=0,n[13]=0,n[14]=0),n[0]=1,n[5]=1,n[10]=1,n[15]=1,n}function q(n,e,t,a,i,r,y,l,s,c,v,u,w,b,h,A){var x=new T(16);return x[0]=n,x[1]=e,x[2]=t,x[3]=a,x[4]=i,x[5]=r,x[6]=y,x[7]=l,x[8]=s,x[9]=c,x[10]=v,x[11]=u,x[12]=w,x[13]=b,x[14]=h,x[15]=A,x}function J(n,e){if(n===e){var t=e[1],a=e[2],i=e[3],r=e[6],y=e[7],l=e[11];n[1]=e[4],n[2]=e[8],n[3]=e[12],n[4]=t,n[6]=e[9],n[7]=e[13],n[8]=a,n[9]=r,n[11]=e[14],n[12]=i,n[13]=y,n[14]=l}else n[0]=e[0],n[1]=e[4],n[2]=e[8],n[3]=e[12],n[4]=e[1],n[5]=e[5],n[6]=e[9],n[7]=e[13],n[8]=e[2],n[9]=e[6],n[10]=e[10],n[11]=e[14],n[12]=e[3],n[13]=e[7],n[14]=e[11],n[15]=e[15];return n}function sn(n,e){var t=e[0],a=e[1],i=e[2],r=e[3],y=e[4],l=e[5],s=e[6],c=e[7],v=e[8],u=e[9],w=e[10],b=e[11],h=e[12],A=e[13],x=e[14],p=e[15],I=t*l-a*y,o=t*s-i*y,f=t*c-r*y,M=a*s-i*l,d=a*c-r*l,P=i*c-r*s,S=v*A-u*h,U=v*x-w*h,F=v*p-b*h,Y=u*x-w*A,g=u*p-b*A,L=w*p-b*x,z=I*L-o*g+f*Y+M*F-d*U+P*S;return z?(z=1/z,n[0]=(l*L-s*g+c*Y)*z,n[1]=(i*g-a*L-r*Y)*z,n[2]=(A*P-x*d+p*M)*z,n[3]=(w*d-u*P-b*M)*z,n[4]=(s*F-y*L-c*U)*z,n[5]=(t*L-i*F+r*U)*z,n[6]=(x*f-h*P-p*o)*z,n[7]=(v*P-w*f+b*o)*z,n[8]=(y*g-l*F+c*S)*z,n[9]=(a*F-t*g-r*S)*z,n[10]=(h*d-A*f+p*I)*z,n[11]=(u*f-v*d-b*I)*z,n[12]=(l*U-y*Y-s*S)*z,n[13]=(t*Y-a*U+i*S)*z,n[14]=(A*o-h*M-x*I)*z,n[15]=(v*M-u*o+w*I)*z,n):null}function wn(n,e,t){var a=e[0],i=e[1],r=e[2],y=e[3],l=e[4],s=e[5],c=e[6],v=e[7],u=e[8],w=e[9],b=e[10],h=e[11],A=e[12],x=e[13],p=e[14],I=e[15],o=t[0],f=t[1],M=t[2],d=t[3];return n[0]=o*a+f*l+M*u+d*A,n[1]=o*i+f*s+M*w+d*x,n[2]=o*r+f*c+M*b+d*p,n[3]=o*y+f*v+M*h+d*I,o=t[4],f=t[5],M=t[6],d=t[7],n[4]=o*a+f*l+M*u+d*A,n[5]=o*i+f*s+M*w+d*x,n[6]=o*r+f*c+M*b+d*p,n[7]=o*y+f*v+M*h+d*I,o=t[8],f=t[9],M=t[10],d=t[11],n[8]=o*a+f*l+M*u+d*A,n[9]=o*i+f*s+M*w+d*x,n[10]=o*r+f*c+M*b+d*p,n[11]=o*y+f*v+M*h+d*I,o=t[12],f=t[13],M=t[14],d=t[15],n[12]=o*a+f*l+M*u+d*A,n[13]=o*i+f*s+M*w+d*x,n[14]=o*r+f*c+M*b+d*p,n[15]=o*y+f*v+M*h+d*I,n}function Mn(n,e,t){var a=t[0],i=t[1],r=t[2];return n[0]=e[0]*a,n[1]=e[1]*a,n[2]=e[2]*a,n[3]=e[3]*a,n[4]=e[4]*i,n[5]=e[5]*i,n[6]=e[6]*i,n[7]=e[7]*i,n[8]=e[8]*r,n[9]=e[9]*r,n[10]=e[10]*r,n[11]=e[11]*r,n[12]=e[12],n[13]=e[13],n[14]=e[14],n[15]=e[15],n}var hn=wn;function on(){var n=new T(3);return T!=Float32Array&&(n[0]=0,n[1]=0,n[2]=0),n}function K(n,e,t){var a=new T(3);return a[0]=n,a[1]=e,a[2]=t,a}function In(n,e,t){return n[0]=e[0]-t[0],n[1]=e[1]-t[1],n[2]=e[2]-t[2],n}(function(){var n=on();return function(e,t,a,i,r,y){var l,s;for(t||(t=3),a||(a=0),i?s=Math.min(i*t+a,e.length):s=e.length,l=a;l<s;l+=t)n[0]=e[l],n[1]=e[l+1],n[2]=e[l+2],r(n,n,y),e[l]=n[0],e[l+1]=n[1],e[l+2]=n[2];return e}})();function cn(){var n=new T(4);return T!=Float32Array&&(n[0]=0,n[1]=0,n[2]=0,n[3]=0),n}function Q(n,e,t,a){var i=new T(4);return i[0]=n,i[1]=e,i[2]=t,i[3]=a,i}function pn(n,e,t){return n[0]=e[0]*t,n[1]=e[1]*t,n[2]=e[2]*t,n[3]=e[3]*t,n}function fn(n,e,t){var a=e[0],i=e[1],r=e[2],y=e[3];return n[0]=t[0]*a+t[4]*i+t[8]*r+t[12]*y,n[1]=t[1]*a+t[5]*i+t[9]*r+t[13]*y,n[2]=t[2]*a+t[6]*i+t[10]*r+t[14]*y,n[3]=t[3]*a+t[7]*i+t[11]*r+t[15]*y,n}(function(){var n=cn();return function(e,t,a,i,r,y){var l,s;for(t||(t=4),a||(a=0),i?s=Math.min(i*t+a,e.length):s=e.length,l=a;l<s;l+=t)n[0]=e[l],n[1]=e[l+1],n[2]=e[l+2],n[3]=e[l+3],r(n,n,y),e[l]=n[0],e[l+1]=n[1],e[l+2]=n[2],e[l+3]=n[3];return e}})();function dn(n,e,t=256,a=1,i=!1){const r=e,y=q(r[0],r[1],r[2],r[3],r[4],r[5],r[6],r[7],r[8],r[9],r[10],r[11],r[12],r[13],r[14],r[15]),l=Q(n[1]/2,n[2]/2,n[3]/2,1),s=cn(),c=B();J(c,y),fn(s,l,c);const v=K(s[0],s[1],s[2]),u=K(a,a,a);let w=q(-1,0,0,0,0,0,1,0,0,-1,0,0,0,0,0,1);i&&(w=q(1,0,0,0,0,1,0,0,0,0,1,0,0,0,0,1)),J(w,w);const b=Q(t,t,t,1),h=B();Mn(h,w,u);const A=Q(b[0],b[1],b[2],1);fn(A,A,h),pn(A,A,.5);const x=on();In(x,v,K(A[0],A[1],A[2]));const p=B();J(p,h),p[3]=x[0],p[7]=x[1],p[11]=x[2];const I=B();sn(I,p);const o=B();hn(o,y,I);const f=B();return sn(f,o),[p,o,f]}function bn(n,e){if(!(n instanceof ArrayBuffer))return n;switch(e){case 2:return new Uint8Array(n);case 4:return new Int16Array(n);case 8:return new Int32Array(n);case 16:return new Float32Array(n);case 64:return new Float64Array(n);case 256:return new Int8Array(n);case 512:return new Uint16Array(n);case 768:return new Uint32Array(n);default:return new Float32Array(n)}}function Fn(n,e,t,a,i){const r=n.length,y=new Uint8Array(r);for(let l=0;l<r;l++){let s=e+i*(n[l]-a);s=Math.max(s,e),s=Math.min(s,t),y[l]=s}return y}function Un(n,e,t,a,i){const r=n.length,y=new Float32Array(r);for(let l=0;l<r;l++){let s=e+i*(n[l]-a);s=Math.max(s,e),s=Math.min(s,t),y[l]=s}return y}function yn(n,e,t,a,i,r,y,l,s,c,v,u=.999){let w=y,b=l;if(n===2)return[w,1];const h=e.length,A=new Float32Array(h);for(let U=0;U<h;U++)A[U]=e[U]*t+a;if(v===0&&u===1)return[w,1];let x=0;for(let U=0;U<h;U++)Math.abs(A[U])>=1e-15&&x++;const p=1e3,I=(b-w)/p,o=new Array(p).fill(0);for(let U=0;U<h;U++){let F=Math.floor((A[U]-w)/I);F=Math.min(F,p-1),o[F]++}const f=new Array(p).fill(0);f[0]=o[0];for(let U=1;U<p;U++)f[U]=f[U-1]+o[U];let M=Math.floor(v*h),d=0;for(;d<p&&!(f[d]>=M);)d++;const P=w;for(w=d*I+P,M=h-Math.floor((1-u)*x),d=0;d<p-1&&!(f[d+1]>=M);)d++;b=d*I+P;let S=1;return w!==b&&(S=(c-s)/(b-w)),[w,S]}function _n(n){const{datatypeCode:e,dims:t,affine:a,sclSlope:i,sclInter:r,toRAS:y=!1,isLinear:l=!0,asFloat32:s=!1,isRobustMinMax:c=!1}=n,v=bn(n.img,e),u=256,w=1,[b,,h]=dn(t,a,u,w,y),A=u*u*u,x=new Float32Array(A),p=t[1]*t[2]*t[3],I=new Float32Array(p);for(let _=0;_<p;_++)I[_]=v[_]*i+r;let o=1/0,f=-1/0;for(let _=0;_<p;_++){const C=I[_];C<o&&(o=C),C>f&&(f=C)}const M=t[1],d=t[2],P=t[3],S=M*d;function U(_,C,j){return _+C*M+j*S}const F=h;let Y=-1;if(l)for(let _=0;_<u;_++)for(let C=0;C<u;C++){const j=C*F[1]+_*F[2]+F[3],en=C*F[5]+_*F[6]+F[7],tn=C*F[9]+_*F[10]+F[11];for(let N=0;N<u;N++){const E=N*F[0]+j,Z=N*F[4]+en,k=N*F[8]+tn,rn=Math.floor(E),ln=Math.floor(Z),an=Math.floor(k);if(Y++,rn<0||ln<0||an<0)continue;const gn=Math.ceil(E),$n=Math.ceil(Z),En=Math.ceil(k);if(gn>=M||$n>=d||En>=P)continue;const O=E-rn,X=Z-ln,m=k-an,D=1-O,G=1-X,H=1-m,$=U(rn,ln,an);let R=0;R+=I[$]*D*G*H,R+=I[$+S]*D*G*m,R+=I[$+M]*D*X*H,R+=I[$+M+S]*D*X*m,R+=I[$+1]*O*G*H,R+=I[$+1+S]*O*G*m,R+=I[$+1+M]*O*X*H,R+=I[$+1+M+S]*O*X*m,x[Y]=R}}else for(let _=0;_<u;_++)for(let C=0;C<u;C++){const j=C*F[1]+_*F[2]+F[3],en=C*F[5]+_*F[6]+F[7],tn=C*F[9]+_*F[10]+F[11];for(let N=0;N<u;N++){const E=Math.round(N*F[0]+j),Z=Math.round(N*F[4]+en),k=Math.round(N*F[8]+tn);Y++,!(E<0||Z<0||k<0)&&(E>=M||Z>=d||k>=P||(x[Y]=I[U(E,Z,k)]))}}const g=c?NaN:0;let L,z,nn;if(s){const[_,C]=yn(e,v,i,r,0,0,o,f,0,1,g);L=Un(x,0,1,_,C),z=16,nn=32}else{const[_,C]=yn(e,v,i,r,0,0,o,f,0,255,g);L=Fn(x,0,255,_,C),z=2,nn=8}const Rn=Array.from(b);return{img:L,dims:[3,u,u,u,1,1,1,1],pixDims:[1,w,w,w,1,0,0,0],affine:Rn,datatypeCode:z,bitsPerVoxel:nn,sclSlope:1,sclInter:0,calMin:0,calMax:0}}function zn(n,e){if(!(n instanceof ArrayBuffer))return n;switch(e){case 2:return new Uint8Array(n);case 4:return new Int16Array(n);case 8:return new Int32Array(n);case 16:return new Float32Array(n);case 64:return new Float64Array(n);case 256:return new Int8Array(n);case 512:return new Uint16Array(n);case 768:return new Uint32Array(n);default:return new Float32Array(n)}}function V(n,e,t,a){return t*a[0]*a[1]+e*a[0]+n}function un(n,e,t){let a=0;const i=new Uint32Array(t+5).fill(0);let y=2147483647;for(let l=0;l<t;l++){let s=e[l];for(a=0;n[s-1]!==s&&(s=n[s-1],a++,!(a>100)););i[l]=s,y=Math.min(y,s)}for(let l=0;l<t;l++)n[i[l]-1]=y}function Cn(n,e,t,a,i,r,y,l){const s=new Uint32Array(27);let c=0;if(!i)return 0;const v=n[V(t,a,i,r)];if(y>=6){const u=V(t,a,i-1,r);v===n[u]&&(s[c++]=e[u])}if(y>=18){if(t){const u=V(t-1,a,i-1,r);v===n[u]&&(s[c++]=e[u])}if(a){const u=V(t,a-1,i-1,r);v===n[u]&&(s[c++]=e[u])}if(t<r[0]-1){const u=V(t+1,a,i-1,r);v===n[u]&&(s[c++]=e[u])}if(a<r[1]-1){const u=V(t,a+1,i-1,r);v===n[u]&&(s[c++]=e[u])}}if(y===26){if(t&&a){const u=V(t-1,a-1,i-1,r);v===n[u]&&(s[c++]=e[u])}if(t<r[0]-1&&a){const u=V(t+1,a-1,i-1,r);v===n[u]&&(s[c++]=e[u])}if(t&&a<r[1]-1){const u=V(t-1,a+1,i-1,r);v===n[u]&&(s[c++]=e[u])}if(t<r[0]-1&&a<r[1]-1){const u=V(t+1,a+1,i-1,r);v===n[u]&&(s[c++]=e[u])}}return c?(un(l,s,c),s[0]):0}function Vn(n,e,t){let a=1;const i=8192;let r=i,y=new Uint32Array(r).fill(0);const l=new Uint32Array(e[0]*e[1]*e[2]).fill(0),s=new Uint32Array(27);for(let c=0;c<e[2];c++)for(let v=0;v<e[1];v++)for(let u=0;u<e[0];u++){let w=0;const b=n[V(u,v,c,e)];if(b!==0){if(s[0]=Cn(n,l,u,v,c,e,t,y),s[0]&&(w+=1),t>=6){if(u){const h=V(u-1,v,c,e);b===n[h]&&(s[w++]=l[h])}if(v){const h=V(u,v-1,c,e);b===n[h]&&(s[w++]=l[h])}}if(t>=18){if(v&&u){const h=V(u-1,v-1,c,e);b===n[h]&&(s[w++]=l[h])}if(v&&u<e[0]-1){const h=V(u+1,v-1,c,e);b===n[h]&&(s[w++]=l[h])}}if(w)l[V(u,v,c,e)]=s[0],un(y,s,w);else{if(l[V(u,v,c,e)]=a,a>=r){r+=i;const h=new Uint32Array(r);h.set(y),y=h}y[a-1]=a,a++}}}for(let c=0;c<a-1;c++){let v=c;for(;y[v]!==v+1;)v=y[v]-1;y[c]=v+1}return[a-1,y,l]}function Sn(n,e,t,a){const i=e[0]*e[1]*e[2];let r=0;const y=new Uint32Array(i).fill(0);for(let c=0;c<a;c++)r=Math.max(r,t[c]);const l=new Uint32Array(r).fill(0);let s=0;for(let c=0;c<i;c++)n[c]&&(l[t[n[c]-1]-1]||(s+=1,l[t[n[c]-1]-1]=s),y[c]=l[t[n[c]-1]-1]);return[s,y]}function Pn(n,e,t){const a=n.length,i=new Uint32Array(e+1).fill(0),r=new Uint32Array(e+1).fill(0);for(let s=0;s<a;s++)i[t[s]]=n[s],r[t[s]]++;let y=0;for(let s=0;s<e+1;s++){const c=i[s];y=Math.max(y,c);for(let v=0;v<e+1;v++)v!==s&&c===i[v]&&(r[s]<r[v]||r[s]===r[v]&&s<v)&&(i[s]=0)}const l=new Uint32Array(a).fill(0);for(let s=0;s<a;s++)l[s]=i[t[s]];return[y,l]}function Nn(n){const{datatypeCode:e,dims:t,conn:a=26,binarize:i=!1,onlyLargestClusterPerClass:r=!1}=n,y=zn(n.img,e),l=Uint32Array.from([t[1],t[2],t[3]]),s=l[0]*l[1]*l[2],c=new Uint32Array(s).fill(0);if(i)for(let f=0;f<s;f++)y[f]!==0&&(c[f]=1);else for(let f=0;f<s;f++)c[f]=y[f];const[v,u,w]=Vn(c,l,a),[b,h]=Sn(w,l,u??new Uint32Array,v);let A,x;r?[A,x]=Pn(c,b,h):(A=b,x=h);let p,I,o;return A>65535?(p=Float32Array.from(x),I=16,o=32):A>255?(p=Uint16Array.from(x),I=512,o=16):(p=Uint8Array.from(x),I=2,o=8),{img:p,datatypeCode:I,bitsPerVoxel:o,calMax:A}}function vn(n,e,t,a,i,r=2){const y=n.length;if(y<1)return[];const l=256,s=l-1,c=new Array(l).fill(0),v=e,u=t;if(u<=v)return[];const w=(u-v)/l;function b(o){return o*w+v}const h=(l-1)/Math.abs(u-v);for(let o=0;o<y;o++){let f=n[o]*i+a;f=Math.min(Math.max(f,v),u),f=Math.round((f-v)*h),c[f]++}const A=Array(l).fill(0).map(()=>Array(l).fill(0)),x=Array(l).fill(0).map(()=>Array(l).fill(0));for(let o=1;o<l;++o)A[o][o]=c[o],x[o][o]=o*c[o];for(let o=1;o<l-1;++o)A[1][o+1]=A[1][o]+c[o+1],x[1][o+1]=x[1][o]+(o+1)*c[o+1];for(let o=2;o<l;o++)for(let f=o+1;f<l;f++)A[o][f]=A[1][f]-A[1][o-1],x[o][f]=x[1][f]-x[1][o-1];for(let o=1;o<l;++o)for(let f=o+1;f<l;f++)A[o][f]!==0&&(A[o][f]=x[o][f]*x[o][f]/A[o][f]);let p=0;const I=[1/0,1/0,1/0];if(r>3)for(let o=0;o<l-3;o++)for(let f=o+1;f<l-2;f++)for(let M=f+1;M<l-1;M++){const d=A[0][o]+A[o+1][f]+A[f+1][M]+A[M+1][s];d>p&&(I[0]=o,I[1]=f,I[2]=M,p=d)}else if(r===3)for(let o=0;o<l-2;o++)for(let f=o+1;f<l-1;f++){const M=A[0][o]+A[o+1][f]+A[f+1][s];M>p&&(I[0]=o,I[1]=f,p=M)}else for(let o=0;o<l-1;o++){const f=A[0][o]+A[o+1][s];f>p&&(I[0]=o,p=f)}return[b(I[0]),b(I[1]),b(I[2])]}function Yn(n,e){const t=n.length,a=new Uint8Array(t);for(let i=0;i<t;i++){const r=n[i];r>e[0]&&(a[i]=1),r>e[1]&&(a[i]=2),r>e[2]&&(a[i]=3)}return a}function Tn(n,e,t,a,i){const r=n.length,y=new Float32Array(r);for(let l=0;l<r;l++){const s=n[l]*t+e;y[l]=s<i?a:n[l]}return y}const W=self.postMessage.bind(self);function xn(n,e){if(!(n instanceof ArrayBuffer))return n;switch(e){case 2:return new Uint8Array(n);case 4:return new Int16Array(n);case 8:return new Int32Array(n);case 16:return new Float32Array(n);case 64:return new Float64Array(n);case 256:return new Int8Array(n);case 512:return new Uint16Array(n);case 768:return new Uint32Array(n);default:return new Float32Array(n)}}function An(n,e,t){let a=Number.POSITIVE_INFINITY,i=Number.NEGATIVE_INFINITY;for(let r=0;r<n.length;r++){const y=n[r]*e+t;Number.isFinite(y)&&(y<a&&(a=y),y>i&&(i=y))}return[a,i]}const Ln={otsu(n,e,t,a,i){const r=i.levels??3,y=xn(n,e),l=t||1,s=a||0,[c,v]=An(y,l,s),u=vn(y,c,v,s,l,r);return{img:Yn(y,u),datatypeCode:2,bitsPerVoxel:8,sclSlope:1,sclInter:0,calMin:0,calMax:r}},removeHaze(n,e,t,a,i){const r=i.level??5,y=r===5||r===1?4:r===4||r===2?3:2,l=xn(n,e),s=t||1,c=a||0,[v,u]=An(l,s,c),w=vn(l,v,u,c,s,y),b=r===1?w[2]:r===2?w[1]:w[0];return{img:Tn(l,c,s,v,b),datatypeCode:16,bitsPerVoxel:32,sclSlope:1,sclInter:0,calMin:v,calMax:u}},conform(n,e,t,a,i){const r=_n({img:n,datatypeCode:e,dims:i.dims,pixDims:i.pixDims,affine:i.affine,sclSlope:t||1,sclInter:a||0,toRAS:i.toRAS??!1,isLinear:i.isLinear??!0,asFloat32:i.asFloat32??!1,isRobustMinMax:i.isRobustMinMax??!1});return{img:r.img,datatypeCode:r.datatypeCode,bitsPerVoxel:r.bitsPerVoxel,sclSlope:r.sclSlope,sclInter:r.sclInter,calMin:r.calMin,calMax:r.calMax,dims:r.dims,pixDims:r.pixDims,affine:r.affine}},connectedLabel(n,e,t,a,i){const r=Nn({img:n,datatypeCode:e,dims:i.dims,conn:i.conn??26,binarize:i.binarize??!1,onlyLargestClusterPerClass:i.onlyLargestClusterPerClass??!1});return{img:r.img,datatypeCode:r.datatypeCode,bitsPerVoxel:r.bitsPerVoxel,sclSlope:1,sclInter:0,calMin:0,calMax:r.calMax}}};self.onmessage=n=>{const{_wbId:e,name:t,img:a,datatypeCode:i,sclSlope:r,sclInter:y,options:l}=n.data,s=Ln[t];if(!s){W({_wbId:e,_wbError:`Unknown transform: ${t}`});return}try{const c=s(a,i,r,y,l??{});W({_wbId:e,...c},[c.img.buffer])}catch(c){W({_wbId:e,_wbError:c instanceof Error?c.message:String(c)})}}})();\n', c = typeof self < "u" && self.Blob && new Blob([f], { type: "text/javascript;charset=utf-8" });
-function y(n) {
-  let e;
-  try {
-    if (e = c && (self.URL || self.webkitURL).createObjectURL(c), !e) throw "";
-    const t = new Worker(e, {
-      name: n == null ? void 0 : n.name
-    });
-    return t.addEventListener("error", () => {
-      (self.URL || self.webkitURL).revokeObjectURL(e);
-    }), t;
-  } catch {
-    return new Worker(
-      "data:text/javascript;charset=utf-8," + encodeURIComponent(f),
-      {
-        name: n == null ? void 0 : n.name
-      }
-    );
-  } finally {
-    e && (self.URL || self.webkitURL).revokeObjectURL(e);
-  }
-}
-let i = null;
-function d() {
-  return i || (i = new u(() => new y())), i;
-}
-async function s(n, e, t, l) {
-  const r = await d().execute({
-    name: n,
-    img: t,
-    datatypeCode: e.datatypeCode,
-    sclSlope: e.scl_slope,
-    sclInter: e.scl_inter,
-    options: l
-  }), a = JSON.parse(JSON.stringify(e));
-  if (a.datatypeCode = r.datatypeCode, a.numBitsPerVoxel = r.bitsPerVoxel, a.scl_slope = r.sclSlope, a.scl_inter = r.sclInter, a.cal_min = r.calMin, a.cal_max = r.calMax, r.dims && (a.dims = r.dims), r.pixDims && (a.pixDims = r.pixDims), r.affine) {
-    const o = r.affine;
-    a.affine = [
-      [o[0], o[1], o[2], o[3]],
-      [o[4], o[5], o[6], o[7]],
-      [o[8], o[9], o[10], o[11]],
-      [0, 0, 0, 1]
-    ], a.sform_code = 1, a.qform_code = 0;
-  }
-  return { hdr: a, img: r.img };
-}
-const w = {
-  name: "otsu",
-  description: "Otsu multi-level thresholding — produces a labeled segmentation volume",
-  options: [
-    {
-      name: "levels",
-      type: "select",
-      label: "Segmentation levels",
-      default: 3,
-      options: [2, 3, 4]
-    }
-  ],
-  resultDefaults: {
-    colormap: "actc",
-    opacity: 0.5
-  },
-  apply: (n, e, t) => s("otsu", n, e, t)
-}, A = {
-  name: "conform",
-  description: "Reslice to 256×256×256 isotropic 1 mm volume (FreeSurfer style)",
-  options: [
-    {
-      name: "toRAS",
-      type: "checkbox",
-      label: "Output RAS orientation",
-      default: !1
-    },
-    {
-      name: "isLinear",
-      type: "checkbox",
-      label: "Linear interpolation",
-      default: !0
-    },
-    {
-      name: "asFloat32",
-      type: "checkbox",
-      label: "Output as Float32",
-      default: !1
-    },
-    {
-      name: "isRobustMinMax",
-      type: "checkbox",
-      label: "Robust min/max (2%-98%)",
-      default: !1
-    }
-  ],
-  apply: (n, e, t) => {
-    const l = {
-      ...t,
-      dims: [...n.dims],
-      pixDims: [...n.pixDims],
-      affine: n.affine.flat()
-    };
-    return s("conform", n, e, l);
-  }
-}, m = {
-  name: "connectedLabel",
-  description: "Label connected components in a volume (bwlabel)",
-  options: [
-    {
-      name: "conn",
-      type: "select",
-      label: "Connectivity",
-      default: 26,
-      options: [6, 18, 26]
-    },
-    {
-      name: "binarize",
-      type: "checkbox",
-      label: "Binarize input",
-      default: !1
-    },
-    {
-      name: "onlyLargestClusterPerClass",
-      type: "checkbox",
-      label: "Only largest cluster per class",
-      default: !1
-    }
-  ],
-  resultDefaults: {
-    colormap: "random",
-    opacity: 0.8
-  },
-  apply: (n, e, t) => {
-    const l = { ...t, dims: [...n.dims] };
-    return s("connectedLabel", n, e, l);
-  }
-}, v = {
-  name: "removeHaze",
-  description: "Remove dark background voxels using Otsu thresholding",
-  options: [
-    {
-      name: "level",
-      type: "select",
-      label: "Preservation level (1=aggressive, 5=conservative)",
-      default: 5,
-      options: [1, 2, 3, 4, 5]
-    }
-  ],
-  apply: (n, e, t) => s("removeHaze", n, e, t)
-};
-export {
-  A as conform,
-  m as connectedLabel,
-  w as otsu,
-  v as removeHaze
-};

package/dist/processing/conform.d.ts

@@ -1,40 +0,0 @@
-/**
- * FreeSurfer-style conform transform — pure computation.
- *
- * Reslices any image to a 256×256×256 isotropic volume with 1 mm voxels,
- * centered on the original volume. Optionally rescales intensities.
- *
- * Based on FastSurfer conform.py (Apache License) and nibabel mghformat.py (MIT License).
- */
-export interface ConformInput {
-    /** Flat voxel data */
-    img: ArrayLike<number>;
-    /** NIfTI datatype code */
-    datatypeCode: number;
-    /** Volume dimensions [ndim, x, y, z, ...] */
-    dims: number[];
-    /** Pixel dimensions [qfac, dx, dy, dz, ...] */
-    pixDims: number[];
-    /** 4×4 affine as flat 16-element array (row-major) */
-    affine: number[];
-    sclSlope: number;
-    sclInter: number;
-    /** Options */
-    toRAS?: boolean;
-    isLinear?: boolean;
-    asFloat32?: boolean;
-    isRobustMinMax?: boolean;
-}
-export interface ConformOutput {
-    img: ArrayBufferView;
-    dims: number[];
-    pixDims: number[];
-    affine: number[];
-    datatypeCode: number;
-    bitsPerVoxel: number;
-    sclSlope: number;
-    sclInter: number;
-    calMin: number;
-    calMax: number;
-}
-export declare function computeConform(input: ConformInput): ConformOutput;

package/dist/processing/connectedLabel.d.ts

@@ -1,20 +0,0 @@
-/**
- * Connected component labeling — pure computation.
- *
- * Port of https://github.com/rordenlab/niimath/blob/master/src/bwlabel.c
- */
-export interface ConnectedLabelInput {
-    img: ArrayLike<number> | ArrayBuffer;
-    datatypeCode: number;
-    dims: number[];
-    conn?: number;
-    binarize?: boolean;
-    onlyLargestClusterPerClass?: boolean;
-}
-export interface ConnectedLabelOutput {
-    img: ArrayBufferView;
-    datatypeCode: number;
-    bitsPerVoxel: number;
-    calMax: number;
-}
-export declare function computeConnectedLabel(input: ConnectedLabelInput): ConnectedLabelOutput;

package/dist/processing/otsu.d.ts

@@ -1,22 +0,0 @@
-/**
- * Pure Otsu thresholding functions.
- * Ported from the old niivue ImageProcessing module.
- *
- * These functions are designed to run in a Web Worker (no DOM dependencies).
- */
-/**
- * Compute one or more Otsu threshold levels for a volume.
- *
- * Based on:
- * - C: https://github.com/rordenlab/niimath
- * - Java: https://github.com/stevenjwest/Multi_OTSU_Segmentation
- */
-export declare function findOtsu(img: ArrayLike<number>, calMin: number, calMax: number, sclInter: number, sclSlope: number, mlevel?: number): number[];
-/**
- * Apply Otsu thresholds to produce a labeled segmentation.
- */
-export declare function applyOtsu(img: ArrayLike<number>, thresholds: number[]): Uint8Array;
-/**
- * Apply haze removal: zero out voxels below the threshold.
- */
-export declare function applyHazeRemoval(img: ArrayLike<number>, sclInter: number, sclSlope: number, globalMin: number, threshold: number): Float32Array;

package/dist/transforms.d.ts

@@ -1,6 +0,0 @@
-import { VolumeTransform } from '@niivue/niivue';
-export declare function terminateWorker(): void;
-export declare const otsu: VolumeTransform;
-export declare const conform: VolumeTransform;
-export declare const connectedLabel: VolumeTransform;
-export declare const removeHaze: VolumeTransform;

package/dist/worker.d.ts

@@ -1,11 +0,0 @@
-/**
- * Web Worker for nv-ext-image-processing transforms.
- *
- * Speaks only primitives — no NIFTI header objects cross the boundary.
- *
- * Protocol (NVWorker bridge):
- *   Request:  { _wbId, name, img, datatypeCode, sclSlope, sclInter, ...options }
- *   Response: { _wbId, img, datatypeCode, bitsPerVoxel, sclSlope, sclInter, calMin, calMax }
- *   Error:    { _wbId, _wbError: string }
- */
-export {};