* feat: add OCR functionality and related configurations

* chore: update labeler configuration for machine learning files

* feat(i18n): enhance OCR model descriptions and add orientation classification and unwarping features

* chore: update Dockerfile to include ccache for improved build performance

* feat(ocr): enhance OCR model configuration with orientation classification and unwarping options, update PaddleOCR integration, and improve response structure

* refactor(ocr): remove OCR_CLEANUP job from enum and type definitions

* refactor(ocr): remove obsolete OCR entity and migration files, and update asset job status and schema to accommodate new OCR table structure

* refactor(ocr): update OCR schema and response structure to use individual coordinates instead of bounding box, and adjust related service and repository files

* feat: enhance OCR configuration and functionality

- Updated OCR settings to include minimum detection box score, minimum detection score, and minimum recognition score.
- Refactored PaddleOCRecognizer to utilize new scoring parameters.
- Introduced new database tables for asset OCR data and search functionality.
- Modified related services and repositories to support the new OCR features.
- Updated translations for improved clarity in settings UI.

* sql changes

* use rapidocr

* change dto

* update web

* update lock

* update api

* store positions as normalized floats

* match column order in db

* update admin ui settings descriptions

fix max resolution key

set min threshold to 0.1

fix bind

* apply config correctly, adjust defaults

* unnecessary model type

* unnecessary sources

* fix(ocr): switch RapidOCR lang type from LangDet to LangRec

* fix(ocr): expose lang_type (LangRec.CH) and font_path on OcrOptions for RapidOCR

* fix(ocr): make OCR text search case- and accent-insensitive using ILIKE + unaccent

* fix(ocr): add OCR search fields

* fix: Add OCR database migration and update ML prediction logic.

* trigrams are already case insensitive

* add tests

* format

* update migrations

* wrong uuid function

* linting

* maybe fix medium tests

* formatting

* fix weblate check

* openapi

* sql

* minor fixes

* maybe fix medium tests part 2

* passing medium tests

* format web

* readd sql

* format dart

* disabled in e2e

* chore: translation ordering

---------

Co-authored-by: mertalev <101130780+mertalev@users.noreply.github.com>
Co-authored-by: Alex Tran <alex.tran1502@gmail.com>
This commit is contained in:
Kang 2025-10-27 22:09:55 +08:00 committed by GitHub
parent c666dc6c67
commit 02b29046b3
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
90 changed files with 3610 additions and 1722 deletions

View file

@ -31,6 +31,7 @@
JobName.VideoConversion,
JobName.StorageTemplateMigration,
JobName.Migration,
JobName.Ocr,
];
// eslint-disable-next-line @typescript-eslint/no-explicit-any

View file

@ -254,6 +254,71 @@
</div>
</SettingAccordion>
<SettingAccordion
key="ocr"
title={$t('admin.machine_learning_ocr')}
subtitle={$t('admin.machine_learning_ocr_description')}
>
<div class="ml-4 mt-4 flex flex-col gap-4">
<SettingSwitch
title={$t('admin.machine_learning_ocr_enabled')}
subtitle={$t('admin.machine_learning_ocr_enabled_description')}
bind:checked={config.machineLearning.ocr.enabled}
disabled={disabled || !config.machineLearning.enabled}
/>
<hr />
<SettingSelect
label={$t('admin.machine_learning_ocr_model')}
desc={$t('admin.machine_learning_ocr_model_description')}
name="ocr-model"
bind:value={config.machineLearning.ocr.modelName}
options={[
{ value: 'PP-OCRv5_server', text: 'PP-OCRv5_server' },
{ value: 'PP-OCRv5_mobile', text: 'PP-OCRv5_mobile' },
]}
disabled={disabled || !config.machineLearning.enabled || !config.machineLearning.ocr.enabled}
isEdited={config.machineLearning.ocr.modelName !== savedConfig.machineLearning.ocr.modelName}
/>
<SettingInputField
inputType={SettingInputFieldType.NUMBER}
label={$t('admin.machine_learning_ocr_min_detection_score')}
description={$t('admin.machine_learning_ocr_min_detection_score_description')}
bind:value={config.machineLearning.ocr.minDetectionScore}
step="0.1"
min={0.1}
max={1}
disabled={disabled || !config.machineLearning.enabled || !config.machineLearning.ocr.enabled}
isEdited={config.machineLearning.ocr.minDetectionScore !==
savedConfig.machineLearning.ocr.minDetectionScore}
/>
<SettingInputField
inputType={SettingInputFieldType.NUMBER}
label={$t('admin.machine_learning_ocr_min_recognition_score')}
description={$t('admin.machine_learning_ocr_min_score_recognition_description')}
bind:value={config.machineLearning.ocr.minRecognitionScore}
step="0.1"
min={0.1}
max={1}
disabled={disabled || !config.machineLearning.enabled || !config.machineLearning.ocr.enabled}
isEdited={config.machineLearning.ocr.minRecognitionScore !==
savedConfig.machineLearning.ocr.minRecognitionScore}
/>
<SettingInputField
inputType={SettingInputFieldType.NUMBER}
label={$t('admin.machine_learning_ocr_max_resolution')}
description={$t('admin.machine_learning_ocr_max_resolution_description')}
bind:value={config.machineLearning.ocr.maxResolution}
min={1}
disabled={disabled || !config.machineLearning.enabled || !config.machineLearning.ocr.enabled}
isEdited={config.machineLearning.ocr.maxResolution !== savedConfig.machineLearning.ocr.maxResolution}
/>
</div>
</SettingAccordion>
<SettingButtonsRow
onReset={(options) => onReset({ ...options, configKeys: ['machineLearning'] })}
onSave={() => onSave({ machineLearning: config.machineLearning })}

View file

@ -19,6 +19,7 @@
mdiTable,
mdiTagFaces,
mdiVideo,
mdiOcr,
} from '@mdi/js';
import type { Component } from 'svelte';
import { t } from 'svelte-i18n';
@ -124,6 +125,14 @@
handleCommand: handleConfirmCommand,
disabled: !$featureFlags.facialRecognition,
},
[JobName.Ocr]: {
icon: mdiOcr,
title: $getJobName(JobName.Ocr),
subtitle: $t('admin.ocr_job_description'),
allText: $t('all'),
missingText: $t('missing'),
disabled: !$featureFlags.ocr,
},
[JobName.VideoConversion]: {
icon: mdiVideo,
title: $getJobName(JobName.VideoConversion),

View file

@ -107,7 +107,7 @@
const onSubmit = () => {
const searchType = getSearchType();
let payload: SmartSearchDto | MetadataSearchDto = {} as SmartSearchDto | MetadataSearchDto;
let payload = {} as SmartSearchDto | MetadataSearchDto;
switch (searchType) {
case 'smart': {
@ -122,6 +122,10 @@
payload = { description: value } as MetadataSearchDto;
break;
}
case 'ocr': {
payload = { ocr: value } as MetadataSearchDto;
break;
}
}
handlePromiseError(handleSearch(payload));
@ -171,17 +175,14 @@
onSubmit();
};
function getSearchType(): 'smart' | 'metadata' | 'description' {
function getSearchType() {
const searchType = localStorage.getItem('searchQueryType');
switch (searchType) {
case 'smart': {
return 'smart';
}
case 'metadata': {
return 'metadata';
}
case 'description': {
return 'description';
case 'smart':
case 'metadata':
case 'description':
case 'ocr': {
return searchType;
}
default: {
return 'smart';
@ -201,6 +202,9 @@
case 'description': {
return $t('description');
}
case 'ocr': {
return $t('ocr');
}
}
}
</script>

View file

@ -4,7 +4,7 @@
interface Props {
query: string | undefined;
queryType?: 'smart' | 'metadata' | 'description';
queryType?: 'smart' | 'metadata' | 'description' | 'ocr';
}
let { query = $bindable(), queryType = $bindable('smart') }: Props = $props();
@ -28,6 +28,7 @@
bind:group={queryType}
value="description"
/>
<RadioButton name="query-type" id="ocr-radio" label={$t('ocr')} bind:group={queryType} value="ocr" />
</div>
</fieldset>
@ -63,4 +64,15 @@
bind:value={query}
aria-labelledby="description-label"
/>
{:else if queryType === 'ocr'}
<label for="ocr-input" class="immich-form-label">{$t('search_by_ocr')}</label>
<input
class="immich-form-input hover:cursor-text w-full !mt-1"
type="text"
id="ocr-input"
name="ocr"
placeholder={$t('search_by_ocr_example')}
bind:value={query}
aria-labelledby="ocr-label"
/>
{/if}